WO2025212971A1 - Gestion de santé panomique intégrative - Google Patents

Gestion de santé panomique intégrative

Info

Publication number
WO2025212971A1
WO2025212971A1 PCT/US2025/023067 US2025023067W WO2025212971A1 WO 2025212971 A1 WO2025212971 A1 WO 2025212971A1 US 2025023067 W US2025023067 W US 2025023067W WO 2025212971 A1 WO2025212971 A1 WO 2025212971A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
subject
health
omic
machine learning
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
PCT/US2025/023067
Other languages
English (en)
Inventor
Warren T. Wamberg
Lauren G. FLANAGAN
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Longevity Life Advisors Inc
Original Assignee
Longevity Life Advisors Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Longevity Life Advisors Inc filed Critical Longevity Life Advisors Inc
Publication of WO2025212971A1 publication Critical patent/WO2025212971A1/fr
Pending legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
    • G16B40/00—ICT specially adapted for biostatistics; ICT specially adapted for bioinformatics-related machine learning or data mining, e.g. knowledge discovery or pattern finding
    • G16B40/20—Supervised data analysis
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
    • G16B20/00—ICT specially adapted for functional genomics or proteomics, e.g. genotype-phenotype associations
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H20/00—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/50—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for simulation or modelling of medical disorders
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients

Definitions

  • said pan-omic analysis is selected by said subject based on one or more preferences.
  • a preference of said one or more preferences is a type of omic data obtained for said subject.
  • said personalized health recommendation comprises a portion of a life insurance product for said subject.
  • said pan- omic analysis cannot be used to negatively impact said subject that is subject to said life insurance product for said subject.
  • said input further comprises an integrated wellness score.
  • said integrated wellness score is determined from a genetic risk score, a health data score, and a lifestyle risk index.
  • said machine learning model is one of a network of machine learning models.
  • said machine learning model is configured to update based on one or more prediction of said network of machine learning models. In further embodiments, said machine learning model is trained to output said personalized health recommendation. In some embodiments, said machine learning model is a large language model (LLM). In further embodiments, said LLM is prompt engineered to provide said personalized health recommendation. In some embodiments, said personalized health recommendation is a lifestyle change or medical intervention. In some embodiments, said personalized health recommendation is used to promote long-term health monitoring.
  • LLM large language model
  • a computer-implemented method of training a machine learning model to provide a personalized health recommendation comprising: providing health data of a first subject to a plurality of trained machine learning models, wherein each trained machine learning model is configured to provide a personalized health recommendation for a subject that is not the first subject; retrieving a plurality of personalized health recommendations from the plurality of trained machine learning models; providing the plurality of personalized health recommendations from the plurality of trained machine learning models to a first machine learning model configured to provide a personalized health recommendation for the first subject; and updating one or more parameters of the first machine learning model.
  • said subject is a policyholder.
  • said health data comprises a pan-omic analysis, medical history data, policy holder data, lifestyle data, or demographic data.
  • said medical history data comprises treatment history, diagnostic history, or medication history.
  • said pan-omic analysis is performed on at least one omic data comprising a genomic data, proteomic data, metabolomic data, epigenomic data, microbiomic data, transcriptomic data, lipidomic data, glycomic data, pharmacogenomic data, nutrigenomic data, phenomic data, or toxicogenomic data.
  • said pan-omic analysis is performed on at least two omic data. In some embodiments, said pan-omic analysis is performed on at least three omic data.
  • said pan-omic analysis is performed on at least four omic data. In some embodiments, said pan-omic analysis is performed on at least five omic data.
  • said omic data comprises a portion that is recommended to said subject. In some embodiments, said at least one of said omic data is selected by said subject.
  • said policy holder data comprises real-time data. In some embodiments, said real- time data comprises health status, treatment efficacy, or system usability. In some embodiments, said health data comprises a bodily sample of said subject. In some embodiments, said health data further comprises a subject preference or a clinician communication.
  • the method further comprises pre-training said machine learning model with a corpus of health data.
  • said machine learning model is an LLM.
  • said plurality of health recommendations comprise a portion of a prompt for use as input into the LLM.
  • said pre-training is a generative pre-training.
  • said machine learning model is a fine-tuned foundation model.
  • at least a portion of said plurality of trained machine learning models are each trained on a subject associated with a trained machine learning model of said plurality of trained machine learning models.
  • at least a portion of said plurality of personalized health recommendations are unique.
  • a parameter of said one or more parameters is a learnable weight of said machine learning model. In some embodiments, a parameter of said one or more parameters is a prompt of said machine learning model. In some embodiments, said personalized health recommendation is used to promote long-term health monitoring.
  • a computer-implemented method of training a machine learning model to provide a personalized health recommendation for a subject comprising: generating health data comprising pan-omic data of a plurality of subjects; performing a generative pre-training training operation for said machine learning model using said health data; generating a recommendation dataset comprising a plurality of patient histories, wherein a patient history of said plurality of patient histories comprises patient health data and patient health recommendation data; and performing a fine-tuning training operation for said machine learning model using said recommendation dataset, wherein said machine learning model is trained to predict said personalized health recommendation for said subject.
  • said health data is de-identified.
  • said health data further comprises one or more of medical history data, policyholder data, lifestyle data, or demographic data.
  • said pan-omic data comprises at least one omic data, wherein omic data comprises a genomic data, proteomic data, metabolomic data, epigenomic data, microbiomic data, transcriptomic data, lipidomic data, glycomic data, pharmacogenomic data, nutrigenomic data, phenomic data, or toxicogenomic data.
  • said pan-omic data comprises at least two omic data.
  • said pan-omic data comprises at least three omic data.
  • said pan-omic data comprises at least four omic data.
  • said pan-omic data comprises at least five omic data.
  • said health dataset further comprises clinician communications.
  • said clinician communications comprises a clinician note to said subject.
  • said clinician communications comprises a conversation between a clinician and said subject.
  • said fine-tuning training operation comprises calculating a probability that said personalized health recommendation is effective.
  • , is a sample of said patient health data, where, is a prior probability that said personalized health recommendation is effective, where,
  • , 1 where, is an aggregation function, where, is a probability of an AI agent of said network that said personalized health recommendation is effective, where, is an innovation function, and where, is a weighting parameter.
  • said personalized health recommendation data comprises a dietary change, a physical activity, a medical consultation, a stress management activity, a medical monitoring, or a mental well-being activity.
  • said machine learning model is a large language model (LLM).
  • said machine learning model is built upon a foundation model.
  • said generative pre-training operation trains said machine learning model to understand a context.
  • said context is said health data.
  • said machine learning model is one of a network of machine learning models.
  • said network of machine learning models is configured to train said network of machine learning models based on said health data.
  • the method further comprises performing a reinforcement learning operation.
  • said reinforcement learning operation comprises reinforcement learning through human feedback (RLHF).
  • said fine-tuning operation a supervised or semi-supervised training operation.
  • a system for generating a personalized health recommendation comprising: one or more computing devices; a subject agent configured to receive a health data from a subject, wherein said subject agent is disposed on a computing device of said one or more computing devices; and a plurality of non-subject agents disposed said computing device, a second computing device of said one or more computing devices, or both, wherein a non-subject agent of said subset of agents is Attorney Docket No.67754-701.601 configured to: receive said health data as input into a machine learning model of said non-subject agent to generate a candidate personalized health recommendation, and provide said candidate personalized health recommendation to said subject agent, wherein said subject agent is further configured to generate said personalized health recommendation based on said health data and said candidate personalized health recommendation.
  • said health data comprises omic data.
  • said omic data comprises genomic data, proteomic data, metabolomic data, epigenomic data, microbiomic data, transcriptomic data, lipidomic data, glycomic data, pharmacogenomic data, nutrigenomic data, phenomic data, or toxicogenomic data.
  • said health data comprises at least two omic data.
  • said health data comprises at least three omic data.
  • said health data comprises at least four omic data.
  • said health data comprises at least five omic data.
  • said health data is encrypted.
  • said subject interacts with said subject agent via a chatbot.
  • said subject agent implements a machine learning model.
  • said subject machine learning model is a large language model (LLM).
  • said machine learning model is trained on said health data.
  • said LLM is prompt-engineered with a prompt based at least in part on said health data.
  • said LLM is trained on de-identified health data of a plurality of subjects associated with said plurality of non-subject agents.
  • a non-subject agent of said plurality of non-subject agents is associated with another subject.
  • said computing device of said one or more computing devices is a subject device.
  • said subject device is a mobile device.
  • said LLM is prompt engineered to provide said personalized health recommendation.
  • said personalized health recommendation is a lifestyle change or medical intervention.
  • said personalized health recommendation is used to promote long-term health monitoring.
  • FIG.1 shows, according to one or more embodiments herein, an example of a computer- implemented system comprising agents.
  • FIG.2 shows, according to one or more embodiments herein, an example of a computing device with one or more processors, memory, storage, and a network interface.
  • FIG.3 shows, according to one or more embodiments herein, an example of a web/mobile application provision system providing browser-based and/or native mobile user interfaces.
  • FIG.4 shows, according to one or more embodiments herein, an example of a cloud-based web/mobile application provision system comprising an elastically load balanced, auto-scaling web server and application server resources as well synchronously replicated databases.
  • FIG.5 shows, according to one or more embodiments herein, an example of the policyholder enrollment. DETAILED DESCRIPTION [0017]
  • the systems and methods disclosed herein may provide proactive, prophylactic, personalized, health recommendations.
  • a life insurance product leveraging such systems and methods may facilitate an insurance ecosystem that empowers life insurance policyholders by prioritizing holistic wellbeing and financial risk mitigation within the life insurance domain.
  • the techniques disclosed herein may enhance medical or insurance equity and protect the privacy of policyholders, while also encouraging an informed, health-aware populace.
  • the systems and methods may provide patient focused health solutions that may improve both pre-diagnostic and post-diagnostic well-being. Additionally, the techniques herein may foster a positive societal impact by enhancing the financial sustainability of life insurance offerings. Further, a policyholder using a life insurance product implementing the systems and methods herein may benefit from a health-focused life insurance experience, which in turn, cultivates a deeper customer-insurer relationship based on mutual interests in health promotion and disease prevention for longevity.
  • the techniques disclosed herein may extend and enhance the lives of policyholders by integrating pan-omics, artificial intelligence, and precision health into personalized health recommendations or insurance policies across a spectrum of health categories such as oncological, cardiovascular, neurological, psychological, gastrointestinal, respiratory, immunological, or other health category.
  • the techniques may promote proactive health management, enabling individuals to mitigate disease risks and optimize wellness outcomes through personalized health recommendations based in part on their unique omic data.
  • Such techniques may supplement or replace extant life insurance policies to provide financial security (e.g., the primary purpose of extant life insurance policies) in addition to actively contributing to the policyholders’ holistic wellbeing and longevity. Implementation of such techniques may foster a deeper, more supportive customer-insurer relationship based on mutual interests in health promotion and disease prevention, setting a new standard for innovation in the insurance industry. Further, the techniques disclosed herein may facilitate the integration of precision health services and products into a network of Attorney Docket No.67754-701.601 agents to streamline access to treatment, time to service, and ease for policyholders to take proactive steps towards continuity of post-diagnostic testing and expedited planning.
  • Network of Agents [0019] Disclosed herein are computer-implemented systems for providing a personalized health recommendation.
  • a computer computer-implemented system 100 for providing a personalized health recommendation may comprise one or more computing devices (e.g., computers, mobile devices, laptops, smart devices).
  • a computing device of the one or more computing devices may be associated with a subject 110.
  • the computing device may host a subject agent 120.
  • the subject agent may comprise one or more machine learning model.
  • the subject agent 120 may comprise subject health data.
  • the subject agent 120 comprise a portion of a network 130 of a plurality of non-subject agents.
  • An agent of the network 130 may be disposed on a computing device of the one or more computing devices.
  • a clinician agent 140 may be associated with the subject agent 120.
  • the clinician agent 140 may be in communication with a clinician 150.
  • the clinician 150 may be selected by, suggested for, or approved by the subject 110.
  • the clinician 150 may be able to interact with subject health data using the clinician agent 140 to provide, approve, alter, reject, or validate a personalized health recommendation provided to the subject 110.
  • An agent of the network 130 e.g., subject agent 120, clinician agent 140, non-subject agent
  • a model of the one or more machine learning models may be trained on health data associated with the subject 110 of the subject agent 120.
  • the subject 110 may be a policyholder of a life insurance product disclosed herein.
  • the subject health data may be used to generate a personalized health recommendation (e.g., lifestyle change, medical intervention, physical activity, a medical consultation, a stress management activity, a medical monitoring, mental well-being activity).
  • a personalized health recommendation e.g., lifestyle change, medical intervention, physical activity, a medical consultation, a stress management activity, a medical monitoring, mental well-being activity.
  • an agent of the network 130 may be registered to another subject in the network 130.
  • the non-subject agents may be able to provide personalized health recommendations based on knowledge of the non-subject agent.
  • a non-subject agent of the network 130 provide a probability that a personalized health recommendation is effective for the subject 110 (e.g., Bayesian updating).
  • one or more model of an agent is fine-tuned or prompted with the health data of a subject.
  • each of a portion of the agents may comprise one or more machine learning models that are fine- tuned or prompted to provide subject specific personalized health recommendations.
  • a subject agent 120 may use the collective knowledge or personalized health recommendations of the network 130 to inform a health strategy of personalized health recommendation for the subject 110 Attorney Docket No.67754-701.601 (e.g., federated learning, centralized learning, in-context learning).
  • the subject agent 110 may aggregate knowledge of the non-subject agents of the network 130 to provide precision health services to the subject 110 that are aware of both health data of the subject 110 and medical knowledge at large (e.g., medical research, clinician communications, non-subject health data).
  • the medical knowledge comprises insights or inputs of the clinician 150 into the clinician agent 140. Medical knowledge may further comprise contemporary treatments, diagnostics, pan-omic analyses, lifestyle interventions (e.g., diet, exercise), or medical research.
  • An agent as disclosed herein may be a part of a network of agents.
  • the network of agents may gather data from a plurality of subjects associated with the network of agents.
  • each agent of the network of agents is associated with a subject.
  • each agent associated with a network of agents is associated with a unique subject.
  • the network of agents may gather health data for a plurality of users using the network. In some embodiments, the gathered health data one or both of de-identified, encrypted, or embedded.
  • a subject agent may be associated with one or more clinician agent.
  • a clinician may comprise a doctor or other medical professional.
  • a clinician may be assigned an agent for monitoring subjects that have approved of the clinician for monitoring health recommendations provided to the subject.
  • a clinician can only view or otherwise interact with subject data if the subject has approved the clinician or opted-in to a policy connected to a plurality of clinicians.
  • Subject health data may comprise subject health goals or clinical interactions.
  • a clinical interaction may comprise voice (e.g., doctor patient conversations), text (e.g., written doctor patient communication), policies (e.g., life insurance policies), or records (e.g., electronic health records, medical records).
  • subject health data may comprise data of a cancer, heart, neuro, life insurance policy, health goals of a subject, or other health documentation (e.g., pan-omic data, medical records).
  • subject health data may comprise a heritable genomic test result, health goals, health/medical records, policy information, pan-omic indicator data, lifestyle information, trackable data (e.g., of a pan-omic analysis), treatment histories, detailed records of past or current treatments, medications, therapies, outcomes.
  • subject 110 may upload or provide real-time data comprising health status, treatment effectiveness, or system usability.
  • the subject 110 health data may comprise omic data.
  • the health data may comprise one or more type of omic data.
  • Types of omic data may comprise genomic data, proteomic data, metabolomic data, epigenomic data, microbiomic data, transcriptomic data, lipidomic data, Attorney Docket No.67754-701.601 glycomic data, pharmacogenomic data, nutrigenomic data, phenomic data, or toxicogenomic data.
  • at least two, at least three, at least four, at least five, at least six, at least seven, at least eight, or more types of omic data may be comprise a portion of the health data of the subject 110.
  • genomic data may indicate a breast cancer risk (e.g., BRCA mutation).
  • This data may be supplemented with transcriptomic data indicating transcription levels or mRNA expression of breast cancer associated genes.
  • the data may be further supplemented with other omic data.
  • a subject 110 with health data including transcriptomic data may provide personalized health recommendations directed to targeted therapies based on the transcriptomic data if the subject 110 is diagnosed with breast cancer.
  • the subject agent 120 may provide personalized health recommendations based on a risk for a condition (e.g., breast cancer) based on health data of the subject (e.g., BRCA mutation).
  • the subject agent 120 may further provide personalized health recommendations based on the transcriptomic data of the subject 110.
  • the combination of pre and post-diagnostic omic data may provide particular utility in health monitoring, early detection, and treatment.
  • the clinician 150 may be provided a treatment option to communicate to the subject 110.
  • the clinician 150 may be notified through the clinician agent 140 that the subject 110 was found to present with a certain pan-omic indicator that has been used to identify effective treatments in the past.
  • precision health solutions to a condition of a subject may be identified through an omics aware machine learning model.
  • the machine learning model may be able provide personalized health recommendations (e.g., therapies, specialists to consult, lifestyle changes) based on one or more pan-omic indicators determined for the subject 110.
  • the condition may be associated with a plurality of health categories comprising cancer, cardiovascular, gastrointestinal, inflammatory neurological, respiratory, immunological, homological, musculoskeletal, psychological, or other health category.
  • an agent e.g., subject agent 120, clinician agent 140
  • the subject health data may comprise a bodily sample of the subject 110.
  • the bodily sample e.g., saliva, blood, urine
  • the pan-omic analysis may result in one or more pan-omic indicator.
  • a pan-omic indicator is a notable aspect of patient health as reported by a laboratory charged with performing the pan-omic analysis or a portion of the pan- Attorney Docket No.67754-701.601 omic analysis.
  • a microbiomic analysis of a pan-omic analysis may be used to indicate subject health.
  • an indicator may be gut bacteria diversity or a population of a gut bacterium.
  • the information may indicate a helpful lifestyle (e.g., diet) change to improve gut health.
  • a genomic analysis of a pan-omic analysis may provide indicators of susceptibility to disease such as breast cancer, Huntington’s disease, or heart disease.
  • a heritable genomic analysis is performed for a subject upon enrollment in the network 130.
  • the heritable genomic analysis may be used to initialize an agent for the subject (e.g., subject agent 120).
  • a plurality of other omic analyses e.g., pan-omic analysis
  • a health data as disclosed herein of a subject at the time of enrollment may be collected for initialization of an agent for the subject.
  • a pan-omic analysis may provide context for the health of the subject 110. Analysis of a single omic data may be informative for simple considerations (e.g., BRCA mutations), but may not provide as comprehensive an understanding as pan-omic analyses for generating personalized health recommendations. In some embodiments, comprehensive understanding may provide particular utility for life insurance policies. Pan-omic analyses may provide deep insight into molecular features of a subject mapped to specific molecular locations (e.g., DNA, RNA, proteins, lipids, glycans) or diagnostically relevant biological features (e.g., phenome, gene expression).
  • Pan- omic analyses may enable the elucidation of valuable insights for targeted treatments across multiple diseases or disease risk factors such that personalized health recommendations may be provided for proactive health or disease management.
  • analysis of pan-omic data may be used in evaluating health data of a subject.
  • Pan-omic data may comprise one or more type of omic data comprising genomic, epigenomic, microbiomic, metabolomic, proteomic, transcriptomic, lipidomic, glycomic, pharmacogenomic, nutrigenomic, toxicogenomic, phenomic, or transcriptomic data.
  • the pan-omic data may be used to provide a personalized health recommendation or as part of a life insurance policy.
  • a subject may undergo comprehensive omic analysis to gain insights into their unique genetic makeup and susceptibility to certain health conditions.
  • comprehensive may indicate at least one, at least two, at least three, at least four, at least five, at least six, at least seven, or more omic data.
  • pan-omic analysis (or a portion thereof) may be done proactively (e.g., upon enrollment in network 130).
  • pan-omic analysis (or a portion thereof) sometimes retroactively (e.g., upon diagnosis).
  • agent only becomes portion of network with opt-in by subject.
  • the systems and methods disclosed herein may provide particular utility in post-diagnostic care for a subject (e.g., subject 110) of the network 130.
  • the subject agent 120 may, upon diagnosis of the subject 110, provide one or more post-diagnostic personalized health recommendations.
  • the personalized health recommendations may be monitoring suggestions (e.g., omic analyses to perform for condition monitoring), lifestyle changes (e.g., diet, exercise change in response to diabetes diagnoses), products or advertisements (e.g., pain management, medications, doctors, clinics, or specialists), or other advice (e.g., as provided by a virtual assistant as disclosed herein).
  • a subject may customize a pan-omic analysis performed for the subject.
  • the subject may customize their pan-omic analysis to aid in risk management.
  • a risk management tool may be a health monitoring (e.g., omic analyses, mental health surveys) or lifestyle change suggestion or action (e.g., exercise, treatment).
  • the subject may be interested in health monitoring due to one or more health concerns.
  • a subject may customize their pan-omic analysis around a specific health concern.
  • the subject may wish to customize their pan-omic analysis on the basis of a known (e.g., medical history) or a perceived (e.g., family history) risk to health.
  • the subject may request a pan-omic analysis for cancer, cardiovascular, neurologic, gastrointestinal, psychological, respiratory, inflammatory, or other health category illnesses.
  • the subject may be able to supplement or request further interpretations over time.
  • a subject receiving personalized health recommendations for cancer and cardiovascular concerns may choose to add neurological interpretations.
  • a subject may decide to alter a plan (e.g., life insurance product) to comprise of a post-diagnostic option only.
  • the subject may choose to receive less proactive screening or health monitoring to reduce their annual premium.
  • genomic data may be analyzed to determine one or more pan-omic indicator.
  • a genomic indicator may include a mutation or sequence of DNA that is known or suspected to contribute to illness.
  • any data or metric gathered about a genetic code may be used as a pan-omic indicator.
  • a presence of a sequence of nucleotides, a mutation in a sequence of nucleotides, an absence of a sequence of nucleotides, or single nucleotide polymorphisms may comprise a pan-omic indicator.
  • proteomic data may be analyzed to determine one or more pan-omic indicator.
  • a proteomic indicator may include data of a protein that is known or suspected to be associated with a health condition (e.g., illness, predisposition, risk).
  • a health condition e.g., illness, predisposition, risk
  • any data or metric gathered about the proteins expressed in a subject may be used as a pan-omic indicator.
  • a protein composition e.g., types of proteins
  • structure, activity, or concentrations of a protein may comprise a pan-omic indicator.
  • a protein interaction with another omic disclosed herein may comprise a pan-omic indicator.
  • metabolomic data may be analyzed to determine one or more pan- omic indicator.
  • a metabolomic indicator may include data of small molecule substrates, metabolites, products, or intermediates of cellular metabolism.
  • any data or metric gathered about the molecular profile in a subject may be used as a pan-omic indicator.
  • a level, concentration, presence, absence, or activity of a metabolite, small molecule, product, or intermediate of cellular metabolism may comprise a pan-omic indicator.
  • epigenomic data may be analyzed to determine one or more pan- omic indicator.
  • An epigenomic indicator may include data of epigenetic mutations of the genetic material of a cell or subject. Generally, any data or metric gathered about the epigenetic modification, activation, or de-activation of genes via chemical medication may be used as a pan- omic indicator. For example, a level of histone modification or DNA methylation may comprise a pan-omic indicator. [0034] In some embodiments, microbiomic data may be analyzed to determine one or more pan- omic indicator. A microbiomic indicator may include data of the microbiome of a subject. Generally, any data or metric gathered about the microbiota including bacteria, archaea, fungi, algae, or small protists may be used as a pan-omic indicator.
  • transcriptomic data may be analyzed to determine one or more pan- omic indicator.
  • a transcriptomic indicator may include data of the transcriptome of the subject.
  • any data or metric gathered about a set of RNA (e.g., RNA, mRNA, miRNA, ncRNA, rRNA, tRNA) transcripts of a subject may be used as a pan-omic indicator.
  • a level of expression, activity, dormancy, or gene regulation may comprise a pan-omic indicator.
  • lipidomic data may be analyzed to determine one or more pan-omic indicator.
  • a lipidomic indicator may include data of lipid molecules in a subject. Generally, any data or metric gathered about lipids in a subject may be used as a pan-omic indicator. For example, Attorney Docket No.67754-701.601 a level, structure, function, dynamic, or interaction of a lipid (e.g., with other molecules) may comprise a pan-omic indicator.
  • glycomic data may be analyzed to determine one or more pan-omic indicator.
  • a glycomic indicator may include data of sugars or glycan structures in a subject.
  • a pharmokinetic, pharmacodynamic, or immunogenic property of a drug in view of a subject’s genes may comprise a pan-omic indicator.
  • nutrigenomic data may be analyzed to determine one or more pan- omic indicator.
  • a nutrigenomic indicator may include data of how nutrition of a subject interact with other omic data of a subject.
  • any data gathered about nutrients of a subject’s interaction with other omics of the subject may be used as a pan-omic indicator.
  • a nutrigenomic indicator may comprise an effect of a nutrition of a subject on the health of the subject.
  • phenomic data may be analyzed to determine one or more pan-omic indicator.
  • a phenomic indicator may include data of the expressed traits of a subject. Generally, any data gathered on the observable characteristics or morphologies of a subject may be used as a pan-omic indicator. For example, a phenomic indicator may be an observable trait of a subject including physical and cognitive traits.
  • toxicogenomic data may be analyzed to determine one or more pan- omic indicator.
  • a toxicogenomic indicator may include data of how another omic data of a subject interacts with a toxin.
  • any data gathered regarding an effect of a toxin on a subject may be used as a pan-omic indicator.
  • a toxicogenomic indicator may be a measure of a susceptibility to or expression of toxins.
  • the subject health data is fed into the network 130 for learning by other agents of the network 130 (e.g., clinician agent 150).
  • the subject data is distributed throughout the network using one or more of embeddings, encryption, or blockchain technology.
  • the subject health data may comprise a pan-omic analysis performed by Attorney Docket No.67754-701.601 one or more laboratory. The results of the pan-omic analysis may be provided to a subject securely to protect subject 110 privacy.
  • the pan-omic analysis results in one or more pan-omic indicator.
  • the one or more pan-omic indicator may be provided to the subject 110 in an encrypted form.
  • An encrypted pan-omic indicator may be de-encrypted on a subject device for evaluation by the subject 110.
  • an encrypted, embedded, or de-identified indicator may be provided to the clinician 150 approved by the subject to review the results of the pan-omic analysis of the subject.
  • the one or more pan-omic indicator may be provided to the subject 110 in an embedded form. An embedded pan- omic indicator may be decoded on a subject device for evaluation by the subject 110.
  • an embedded indicator may be provided to a clinician 150 approved by the subject 110 to review the results of the pan-omic analysis of the subject.
  • the embedded representations are an output of a machine learning model (e.g., encoder, BERT).
  • An encrypted or embedded representation of a pan-omic indicator may be unintelligible without an appropriate decryption or decoder algorithm. This may improve data security during transfer of subject data (e.g., over a network, among agents).
  • encryption or embedding may be used for any aspect of subject health data as disclosed herein.
  • an agent comprises one or more machine learning model.
  • the machine learning model may be fine-tuned (e.g., supervised learning) to perform a task based on training data.
  • a machine learning model may be fine-tuned and prompted-engineered to provide personalized health recommendations.
  • subject health data may be de-identified before being used in training a non- subject agent of the network 130.
  • the network 130 may comprise a plurality of agents associated with subjects or clinicians. The agents of the network 130 may each gather valuable health data of a subject (e.g., subject, clinician, nurse) associated with an agent of the network 130.
  • Attorney Docket No.67754-701.601 subjects may provide subject health data, prior life insurance policy data, medical records, or communications (e.g., with clinicians, an agent).
  • a clinician e.g., clinician 150
  • the data gathered by the network 130 may be used to train machine learning models associated with agents of the network 130 (e.g., federated learning, centralized learning). The data gathered by the network 130 may be used to continuously improve robustness of agents of the network 130.
  • the virtual assistant may be a chat bot.
  • the chat bot may be powered by an LLM.
  • the virtual assistant may provide or consider a personalized health recommendation to enhance customer engagement or satisfaction by providing convenient access to information, guidance, and support.
  • the virtual assistant may be interactive. Interaction with the virtual assistant may be via a large language model or LLM (e.g., ChatGPT).
  • LLM e.g., ChatGPT
  • An LLM implementation of a virtual assistant may be prompted to provide more information about a personalized health recommendation or life insurance product.
  • the subject 110 may request more information about how to implement a lifestyle change. Following the example, a dietary change lifestyle change may result in a request for “high fiber food suggestions” by the subject.
  • Information may comprise clarification, further explanation, or other information necessary for the subject 110 to implement a personalized health recommendation.
  • the LLM implementation of a virtual system may be prompted to provide guidance to the subject.
  • the subject may request advice for carrying out a personalized health recommendation.
  • Guidance may comprise steps to take to implement a personalized health recommendation or a motivation behind a personalized health recommendation.
  • Attorney Docket No.67754-701.601 the virtual assistant may advise the subject 110 to receive regular blood pressure monitoring due to indicator from a pan-omic analysis of the subject 110 indicating an elevated risk of heart disease.
  • the LLM implementation of a virtual assistant may provide support to the subject 110.
  • the virtual assistant my provide recommendations of specialists or health providers to consult regarding a personalized health recommendation.
  • Supporting the subject may comprise following-up with the subject regarding progress on personalized health recommendations or lifestyle changes motivated by personalized health recommendations.
  • Support may comprise suggestions for medical interventions such as diagnostics to seek or omic analyses to have performed.
  • the virtual assistant may advise subject 110 to obtain an updated microbiomic analysis following a recommended diet change.
  • Support may comprise putting the subject 110 in contact with a human counterpart to prepare for treatment should a disease or diagnosis occur.
  • subjects may interact with a virtual assistant to address inquiries, receive health advice, and manage their integrative pan-omic life insurance policies more efficiently.
  • Virtual assistants my provide particular utility in patient adherence to health monitoring, lifestyle changes, or treatments. This may lead to a better overall customer experience.
  • management general agencies or brokerage firms may be assigned an agent to support a subject. In some embodiments, management general agencies or brokerage firms may provide financial support or guidance through their associated agents to the subject 110 or other subjects of the network 130. For example, management general agencies may facilitate a subject in enrolling in a life insurance product incorporating the systems and methods disclosed herein.
  • the network 130 as disclosed herein may be used to generate documentation or reports demonstrating the network’s compliance with data privacy standards and regulations. In some embodiments, compliance with data privacy standards may comprise subject opt-in or data de-identification, encryption, or embedding. The network 130 may generate updates or adjustments to machine learning models of agents based on new pan-omic data, shared learnings, or feedback.
  • the network may enable the monitoring of health data integration rate (e.g., into training), health recommendation integration rate (e.g., subject adherence or clinician personalized health recommendation adoption rate).
  • health recommendation integration rate e.g., subject adherence or clinician personalized health recommendation adoption rate
  • the network may provide a measure of the integration of pan-omic data, combined with detailed health goals, or real-time clinician input/data. This may enable agents to generate highly personalized treatment recommendations (e.g., for health monitoring, treatment, preventative care).
  • personalized health recommendations provided to the subject 110 are provided, approved, validated, edited, or otherwise monitored by the clinician 110 or another approved clinician.
  • data used for training a machine learning model herein e.g., a model of the subject agent 120, a model of the clinician agent 140, a model of a non-subject agent of the network 130, or other model referenced herein
  • the machine learning models disclosed herein may provide highly robust, contemporarily relevant, clinically sound personalized health recommendations.
  • a model of the network 130 may be initialized with clinically validated data prior to fine-tuning on subject health data.
  • a pre-trained model may be deployed to agents of the network 130 after training on clinically validated data.
  • Clinically validated data may comprise one or both of labelled (e.g., for supervised, semi-supervised) or unlabeled (e.g., for unsupervised, self-supervised) data that has been provided, approved, validated, edited, or otherwise monitored by the clinician 110 or another approved clinician prior to training. This may provide particular utility in providing high quality, reliable training data from which machine learning models of the agents of the network 130 may learn.
  • one or more machine learning models of an agent of the network 130 may function as a nurse navigator. A nurse navigator may be trained in omics and patient communication.
  • the nurse navigator may comprise a portion of a virtual assistant as disclosed herein.
  • the nurse navigator may confer with a subject to discuss findings (e.g., relevance of pan-omic indicators) to their health.
  • the nurse navigator may comprise a subject-facing portion of an agent (e.g., subject agent 120, non-subject agent, clinician agent 140) that is used for interacting with the subject (e.g., guidance, support).
  • an agent e.g., subject agent 120, non-subject agent, clinician agent 140
  • a nurse navigator may be used review a genomic, microbiomic, transcriptomic, or other omic analysis result.
  • the nurse navigator may be trained as a fine-tuned LLM to be contextually away of the relevance of pan-omic analyses as disclosed herein.
  • the contextual awareness of the nurse navigator may be derived from training a model of the nurse navigator on clinician communications or other medical knowledge to provide data regarding effective communication of Attorney Docket No.67754-701.601 pan-omic indicator relevance.
  • communication of the nurse navigator may be recorded and saved for record-keeping or pre-training (e.g., federated learning, centralized learning).
  • the nurse navigator may be particularly suited to explaining pan-omic analysis results or significance, potential health goals or strategies are also tailored to personal health preferences of the subject 110.
  • health data may be provided directly from subject pan-omic test results, medical documents, medical test results, governmental regulatory rules and guidelines, federal/state/local regulatory and compliance guidelines, health and insurance regulations, health association guidelines (i.e., ACA, AHA, ACS, etc.), interpretations by clinicians, real-time data (text, chat, voice between clinician and subject), or interpretations made by an agent of a network of agents.
  • health association guidelines i.e., ACA, AHA, ACS, etc.
  • interpretations by clinicians e.e., ACA, AHA, ACS, etc.
  • privacy and security of network 130 subjects may comprise data de- identification.
  • an agent may be configured to perform encryption or anonymization through the use of blockchain as a data transfer technique.
  • the agent may be configured to adhere to regulatory standards.
  • the privacy and security of the network may provide particular utility in compliance and persistence of subjects with the network.
  • interactions among subjects, their agents, testimonials, clinicians, service providers, nurse navigators, and the chosen services/products provides valuable data for continuous improvement of the offerings for pre-training.
  • This de-centralized feedback loop can identify trends, preferences, and outcomes that can guide future enhancements to the centralized network.
  • the network and the agents and machine learning models therein may continuously update to improve the network for precision health.
  • the network 1130 may apply Talagrand’s inequalities when the network 130 transfers a pre-trained agent to a post-diagnostic pan-omic technology partner (e.g., laboratory for performing a portion of a pan- omic analysis).
  • a stochastic system for pre-training an agent may provide particular utility as the network grows.
  • Talagrand’s inequalities may be used to analyze and secure the distribution and behavior of the agents as they move in and out of the network 130.
  • Talagrand’s inequalities may help in understanding how the collective behavior of agents of the network may deviate from expected outcomes and mitigate potential risks and security threats.
  • An agent as disclosed herein may be provided to provide a personalized health recommendation.
  • Providing the personalized health recommendation may comprise operations including (1) obtaining a bodily sample from a subject, (2) performing a pan-omic analysis to obtain one or more pan-omic indicator using the bodily sample from the subject, (3) determining the personalized health recommendation for the subject using an input into the machine learning model trained on health data of the person, and (4) providing the personalized health recommendation to the person.
  • the agent may use one or more machine learning models to perform a portion of the operations provide the personalized health recommendation.
  • the personalized health recommendation may comprise a lifestyle change (e.g., change to diet or exercise) or medical intervention (e.g., doctor or specialist to consult, omic analysis to have performed, diagnostic to seek).
  • the bodily sample from the subject may comprise a saliva, urine, fecal, blood, hair, mucus, cheek swab, or lymph sample.
  • the sample may be provided by the subject may be collected at home or otherwise by the subject. Alternatively, or additionally, the bodily sample may be collected by a clinician.
  • the bodily sample may comprise one or more of saliva, urine, fecal, blood, hair, sweat, mucus, cheek swab, or lymph and be collected by one or both of the subject or a clinician.
  • the bodily sample may comprise a saliva sample or fecal sample collected by a subject at home.
  • the subject may provide the saliva sample while a clinician draws a blood sample from the subject.
  • the bodily sample may be collected by the subject at home (e.g., not at a clinic, medical office) for sending to one or more laboratory.
  • the bodily of the sample may be analyzed in a pan-omic analysis by the one or more laboratories.
  • a laboratory for performing genomic analysis may perform a portion of the pan-omic analysis while another laboratory performs a nutrigenomic analysis of the pan-omic analysis.
  • a lab of the one or more laboratories may analyze the bodily sample and determine one or more pan-omic indicator.
  • the pan-omic indicator may comprise a health metric that has been shown to be associated with a health condition. For example, statistical analysis has shown a BRCA gene mutation to be associated with breast cancer risk. Further, a transcriptomic laboratory may provide an indicator of a health concern (e.g., imminent, nascent, eventual) based on a level of gene expression measured in the subject. Yet further, a proteomic analysis may indicate cardiovascular health based on levels of proteins in blood (e.g., growth differentiation factor, GDF). The composition of a pan-omic analysis may be personalized by a subject (e.g., preference) or for a subject (e.g., clinician or agent suggestion).
  • a subject e.g., preference
  • a subject e.g., clinician or agent suggestion
  • An omic analysis comprising a portion of a pan-omic analysis may provide one or more pan-omic indicator.
  • a pan-omic indicator may be a result of an omic analysis that indicates a certain health risk.
  • the health risk may be indicated by a result of the omic analysis that may be Attorney Docket No.67754-701.601 abnormal as compared to a population of other subjects subjected to the same omic analysis.
  • high baseline levels of blood insulin versus a population average may be a pan-omic indicator of potential health concern.
  • one or more omic analysis may be performed for the subject upon enrollment in a network of agents.
  • one or more omic analysis may be performed in response to a diagnosis. In some embodiments, one or more omic analysis may be performed more than once over time for health monitoring. In some embodiments, one or more omic analysis may be requested by the subject due to their preference or personal health concerns. For example, a personalized health plan for the subject may enable the subject to select a pan-omic analysis to be performed for the subject. Furthering the example, a subject may have greater concern with cardiovascular health versus other categories of health such as cancer risk or neurological disease risk. The subject may indicate this preference to an agent associated with the subject and ultimately receive the information (e.g., at home test kit, advertisement for appropriate doctor or service provider) necessary to establish a personalized health plan based on the subject preference.
  • information e.g., at home test kit, advertisement for appropriate doctor or service provider
  • the subject may indicate preference to the agent via manual selection (e.g., via web application) or via conversation with the agent associated with the subject (e.g., chatbot, virtual assistant).
  • one or more omic analysis may be suggested to the subject by an agent associated with the subject.
  • a clinician may suggest one or more omic analysis be performed for a subject based on a suggestion of an agent associated with the clinician.
  • the health data of the subject may comprise one or more pan-omic indicator, medical history of the subject, or demographic data of the subject.
  • a calculation may be performed on a genetic risk (e.g., genetic risk score), a transcriptomic risk (e.g., transcriptomic risk score), a nutrigenomic risk (e.g., nutrigenomic risk score), or any other type of risk based on omic types as disclosed herein.
  • the integrated wellness score may comprise a health data score.
  • the health data score may comprise a calculation based on risk associated with health data of a subject. For example, the health data score may consider family history, medical history, existing diagnoses, medications, pre-existing conditions, allergies, mental health, or other health data of the subject.
  • the integrated wellness score may comprise a lifestyle risk index.
  • the lifestyle risk index may comprise diet, exercise, stress levels, blood pressure, or other lifestyle indicator of the subject.
  • genomic indicators may be used to identify a certainty of onset of a condition (e.g., Huntington’s disease) or risk for a condition (e.g., Parkinson’s disease).
  • micrbiomic indicators may indicate a level of gut health or biodiversity in the flora of a subject microbiota.
  • transcriptomic data may provide an indicator of gene expression in a subject (e.g., elevated, normal, or decreased gene expression resultant from disease or therapy).
  • the aggregation Attorney Docket No.67754-701.601 of pan-omic indicators from a pan-omic analysis may provide comprehensive into subject health from the genetic and molecular level to the phenotype level.
  • the subject health data is de-identified (e.g., encrypted, embedded) prior for use in training.
  • the subject health dataset may comprise medical history (e.g., prior treatments, illness, surgery, medications, allergies, family history), subject data (e.g., preferences, health concerns, omic analyses desired in pan-omic analysis), lifestyle data (e.g., diet, activity levels), or demographic data (e.g., ancestry, geographic location).
  • a pan-omic analysis may include analyses of a number of omic data types.
  • Omic data types may comprise genomic data, proteomic data, metabolomic data, epigenomic data, microbiomic data, transcriptomic data, lipidomic data, glycomic data, pharmacogenomic data, nutrigenomic data, phenomic data, or toxicogenomic data.
  • the pan-omic analysis comprises at least one, at least two, at least three, at least four, at least five, at least six, at least seven, at least eight, or more types of omic data.
  • Each type of omic data evaluated in a pan-omic analysis may provide one or more pan-omic indicator for incorporation into subject health data.
  • the subject health data may comprise patient communications.
  • Patient communications may comprise patient (e.g., subject) doctor communications about lifestyle changes, possible treatment options, diagnoses, prognoses, or other health information.
  • This data may provide domain or context information to the machine learning model.
  • a machine learning model may learn the contextual relevance of pan-omic indicators by including patient communications. This may provide particular utility in pre-training operations that may be used to fine-tune a foundation model (e.g., GPT, LLaMa, PaLM).
  • a doctor note to a patient regarding the relevance of findings of an omic or pan-omic analysis may facilitate domain learning by the machine learning model.
  • the patient communication may comprise a conversation between a clinician and subject.
  • the inclusion of conversations may, in addition to providing enhanced domain learning, be used to train a machine learning model to perform virtual assistant or chatbot operations.
  • a machine learning model e.g., LLM
  • the machine learning techniques disclosed herein may be implement Bayesian updating. Bayesian updating may be used by machine learning models to incorporate learnings of a network of agents to update the network of agents.
  • each agent of the network may comprise one or more machine learning models.
  • at least one of the one or more machine learning models is a large language model.
  • the learnings of the agents of the network may be aggregated to update an agent for a subject.
  • This may Attorney Docket No.67754-701.601 facilitate updating, pre-training, or fine-tuning agents of the network to incorporate updated medical knowledge.
  • the aggregation of learning may provide particular utility in updating machine learning models of the network with the most recent medical knowledge, research, or treatments.
  • agents of the network may be associated with or in communication with clinicians that may validate or provide medical expertise directly to the network or via communications with a subject associated with the network.
  • the Bayesian updating may incorporate a collaborative Bayesian learning network (CLBN) mechanism.
  • CLBN collaborative Bayesian learning network
  • the (CLBN) may implement a probability for updating a belief of an agent.
  • updating a belief of the agent may comprise updating (e.g., training, prompting) a machine learning model of a subject agent.
  • the probability calculated may indicate a probability that a personalized health recommendation is effective for a subject given subject health data.
  • the probability may be described by the probability given by the equation , .
  • the probability ( ) may be a sample of subject (e.g., subject, subject) data.
  • the term may be a prior probability that a particular personalized health recommendation ( ) is effective (or true).
  • the prior probability that the personalized health recommendation is effective may be based on one or both of pre-existing data (e.g., subject health data, non-subject health data) or general medical knowledge.
  • the , term may be a likelihood of observing the subject health data given that the personalized health recommendation is effective.
  • the , term may be a probability of observing the subject health data given a plurality of possible observations derived from the subject health data.
  • the , may be used for normalizing the probability or probabilities of ( ).
  • the term may be a collaborative updating function for a machine learning model of the subject agent.
  • the term may be a model of the subject agent.
  • the term may be a network of the machine learning model (e.g., non-subject agents or a model of non-subject agents, clinician agents).
  • The may be de-identified health data of at least the subject.
  • the may be aggregated health data.
  • the aggregated health data may comprise subject health data or non-subject health data.
  • the innovation function may be described by
  • The may be an aggregation function.
  • the aggregation function may utilize reinforcement, centralized, or federated learning as described herein.
  • the aggregation function may gather data of a plurality of subjects of a network or training information (e.g., updated model weights) of machine learning models of agents of the network.
  • the function may leverage the collective intelligence or learning of the network.
  • term may be a probability of a machine Attorney Docket No.67754-701.601 learning model associated with a non-subject or non-subject agent of the network that the personalized health recommendation is effective.
  • the aggregation function may be used to obtain, for example, an average probability that the personalized health recommendation is effective based on the learnings of other agents of the network.
  • the agents of the network may have similar foundational beliefs (e.g., trained on a shared dataset) that is updated with domain beliefs (e.g., data of a subject associated with an agent) such that the agents may differ on a likely effectiveness of a given personalized health recommendation.
  • This may provide particular utility in a number of circumstances where a case is complex or unique and the opinions of numerous agents (e.g., agents) is desired. It may be considered that the outputs of the agents are reflective of reliable medical knowledge as the agents may be trained on clinician validated data.
  • a subject may have unusual health that biases an agent associated with the subject away from standard personalized health recommendations. By considering the probability that a personalized health recommendation is effective from other sources, the subject agent is protected from possible biases of the subject agent. This may address a critical problem of large language models, which may be prone to assuming that input data is contextually sufficient to resolve inputs.
  • the consideration of a network of beliefs may reduce the risk of a large language model implemented as a portion of an agent from providing low likelihood of success for likely successful personalized health recommendations.
  • the term may be an innovation function.
  • the innovation function may comprise a pre-training operation for training a subject agent on the subject health data.
  • the innovation function may integrate unique or new insights derived from . This may enable to learn from both the subject health data and the health data of other non-subject members of the network.
  • the term may be a weighting parameter. In some embodiments, the weighting parameter may balance the influence of network- derived intelligence versus insights from the subject data.
  • weighting parameter is dependent on a confidence in a network’s collective knowledge (e.g., a quality of training) versus the uniqueness of a subject’s individual data. For example, a low confidence network or highly unique subject data may indicate that a machine learning model of an agent of the subject should prioritize the subject health data in making a personalized health recommendation. In another example, the weighting parameter may be used to control a degree of updating in a machine learning model of a subject agent if the innovation function comprises a training operation. [0071] In some embodiments, Bayesian updating may be based at least in part on graph theory. For example, agents of a network (e.g., network 130) may be organized according to similarity of subjects in the network.
  • subjects may be organized according to clinicians (e.g., subject agents are organized around a mutual clinician’s agent).
  • Organizing the agents of the Attorney Docket No.67754-701.601 network according to graph theory may facilitate the use of highly efficient algorithms for searching, accessing, aggregating, or sorting subjects in a network for collaboration, training, or predictions disclosed herein.
  • Bayesian updating may consider the probabilities of the nearest agents (e.g., as found by a breadth first search) in providing a personalized health recommendation to a subject.
  • the nearest agent may be determined by, for example, a patient similarity (e.g., medical history, current health concern, similarity of health data).
  • an operation of training a machine learning model to provide a personalized health recommendation may comprise performing a fine-tuning training operation for the machine learning model using said recommendation dataset.
  • the machine learning model may be trained to predict a personalized health recommendation for a subject.
  • Fine-tuning the machine learning model may comprise the use of a labelled (e.g., supervised learning) or partially labelled (e.g., semi-supervised learning) dataset.
  • the dataset used to train the machine learning model may comprise clinician approved personalized health recommendations.
  • a dataset may comprise health data of a number of subjects.
  • a clinician may validate, insert, reject, or supplement a personalized health recommendation of associated with the subject health data.
  • an operation of training a machine learning model to provide a personalized health recommendation may comprise performing a reinforcement learning operation.
  • reinforcement learning operation comprises reinforcement learning through human feedback (RLHF).
  • RLHF human feedback
  • a personalized health recommendation may be approved or otherwise rated by a subject upon receipt of the personalized health recommendation.
  • a clinician may provide feedback (e.g., approval, editing, rejection, validation, supplementation) regarding a proposed personalized health recommendation by an agent of the Attorney Docket No.67754-701.601 network. Such feedback may be collected and used to update a machine learning model for quality, robustness, accuracy, or preference accommodation.
  • training a machine learning to model to provide a personalized health recommendation may comprise operations including (1) providing health data of a first subject to a plurality of trained machine learning models, (2) retrieving a plurality of personalized health recommendations from the plurality of trained machine learning models, (3) providing the plurality of personalized health recommendations from the plurality of trained machine learning models to a first machine learning model configured to provide a personalized health recommendation for the first subject , and (4) updating one or more parameters of the first machine learning model.
  • the health data of the subject may comprise pan-omic analysis data (e.g., pan-omic indicator), lifestyle data, medical records, demographic information, subject data, treatment history, medication history, diagnostic history, or other health data component as disclosed herein.
  • the health data in some cases in a de-identified, encrypted, or embedded form, may be provided to one or more machine learning models (e.g., as used in agents of network 130).
  • each of the one or more machine learning models may be associated with another subject.
  • the one or more machine learning models may be used to provide insight or predictions into a proposed personalized health recommendation for the subject given learnings of machine learning models not associated directly with the subject.
  • subject data may be used to train a machine learning model of a non-subject, but a machine learning model associated with the non-subject may be fine-tuned on the health data of the non-subject.
  • the non- subject machine learning model may be trained or prompted on clinician provided data or communications not directly accessible by the subject machine learning model. This may configure a non-subject machine learning model to operate in a different context (e.g., the health data of the non-subject).
  • the non-subject data may have different personalized health recommendations for the same data (e.g., subject data) than the machine learning model (e.g., of an agent) associated with the subject.
  • the English aware LLM may be trained on health data (e.g., medical records, pan- omic analyses, lifestyle data) as a pre-training operation before fine-tuning for specific operations (e.g., providing personalized health recommendations).
  • training may update one or more parameters of the machine learning model.
  • a parameter may be a learnable weight of the machine learning model.
  • health data of non-subject may be used for training a model of an agent as disclosed herein (e.g., centralized learning, federated learning).
  • a parameter may be a prompt of the machine learning model.
  • non-subject health data may be used to generate a prompt for in-context learning of a machine learning model of a subject.
  • a machine learning model e.g., subject agent
  • the agent may be apprised (e.g., by training, prompting) of subject health data such that the agent can provide personalized health recommendations.
  • personalized health recommendations may be provided in real-time. For example, if the subject interacts with an agent (e.g., uploading more data, engaging with virtual assistant), then the agent may update its beliefs through training or prompt updates to provide updated personalized health recommendations.
  • an agent may provide initial personalized health recommendations based on an existing set of beliefs (e.g., health data).
  • the agent may provide updated personalized health recommendations based on one or both of updated subject health data or aggregated beliefs of a network comprising the agent.
  • Large Language Models may comprise from-scratch models trained on a corpus of health data (e.g., medical records, clinician communications, pan-omic analyses). Alternatively, LLMs may build upon a foundation model such that an LLM as disclosed Attorney Docket No.67754-701.601 herein may be trained using a corpus of context specific data using the weights of a foundation model for initializing of the context specific task.
  • Context specific data may comprise labelled (e.g., for supervised or semi-supervised learning) or unlabeled (e.g., for unsupervised or self-supervised learning) data depending on a training goal or goals.
  • a training may comprise a single or multi-task learning.
  • the context specific data may comprise subject health data, pan-omic data, or clinician communication data.
  • Foundation models may provide a base layer of knowledge such that the base layer of knowledge may be supplemented with further training operations comprising unsupervised, supervised, semi-supervised, or self-supervised training.
  • foundation models may comprise Azure OpenAI (complete and chat APIs for GPT-3, GPT-3.5, and GPT-4, used in ChatGPT), OpenAI (complete and chat APIs for GPT-3, GPT-3.5 and GPT-4), Google Vertex AI (e.g., PaLM, PaLM-2), Meta's LLaMa, as well as BLOOM, Ernie 3.0 Titan, and Anthropic's Claude 2, FLAN-T5, OpenAssistant, RoBERTa, MiniLM, MPNet, DALL- E, and BERT. While foundation models may comprise those listed previously, foundation models may generally comprise any transformer or self-attention based model that has been previously trained. [0080] In some embodiments, LLMs may comprise transformers.
  • Transformers may be applied to a variety of natural language processing tasks such as text classification, named entity recognition, summarization, translation question answering, text generation, personalized health recommendation, or virtual assistance (e.g., chatbots). Transformers may additionally be applied to other data modalities as with vision transformers (ViT). Transformers may be enabled by the self- attention mechanism.
  • the self-attention mechanism may enable models implementing the self- attention mechanism to assign different weight to different portions of a sequence of an input. Self- attention may enable LLMs or transformer based models to leverage contextual information of an input to inform an output while reducing the risk of the model “forgetting” about earlier portions of an input.
  • LLMs as disclosed herein may be configured to perform operations that an LLM may or may not have been explicitly trained on. Configuring LLMs to perform operations or resolve inputs on which they were not explicitly trained may comprise prompt-engineering. In some embodiments, the ability of an LLM to perform a task on which it was not explicitly trained may be considered a form of zero-shot learning.
  • a prompt into an LLM may comprise health data and Attorney Docket No.67754-701.601 instructions for how to interact with health data (e.g., to provide a personalized health recommendation).
  • the prompt may comprise a natural language input.
  • a natural language input may indicate that the input is not required to adhere to any specific format and that prompt may be arbitrarily structured for a given task.
  • a prompt may comprise one or more of a textual input, an image input, an audio input, a video input, or a tabular input for use in indicating a desired output of the LLM.
  • the input may comprise an imperative, an interrogative, an exclamatory, or a declarative input.
  • the prompt may comprise any arbitrary component that the LLM is capable of encoding into an embedding that may be compared against a learned latent space of the LLM.
  • a prompt in the context of LLMs may comprise instructions. Instructions may comprise one or more operations to execute. In some embodiments, the operations may be ordered. Instructions may be used to guide an LLM to resolve an input into the LLM. In some embodiments, instructions may be applied to resolve complex or multi-stage problems. In some embodiments, the input may comprise formatting, problem-solving, or organizational preferences. For example, an LLM may be prompt engineered to provide outputs aligned with a custom health concern of a subject (e.g., cardiovascular, cancer, neurodegenerative).
  • a custom health concern of a subject e.g., cardiovascular, cancer, neurodegenerative
  • the prompt may facilitate in-context or few-shot learning.
  • in-context or few-shot learning a small number of examples may be provided to an LLM in the body of a prompt.
  • the examples provided in the prompt may facilitate the LLM in predicting a never before resolved example.
  • subject health data may be analyzed via a prompt comprising insights (e.g., personalized health recommendations) other a plurality of agents of a network of agents.
  • insights e.g., personalized health recommendations
  • the examples contained in the prompt may not be persistently stored in the LLM (e.g., not used to adjust model weights or other parameters).
  • soft prompting may be used in prompting an LLM to resolve a subject input.
  • Soft prompting may provide an efficient route towards leveraging an LLMs ability to perform tasks when given instructions or guidance in the form of a prompt. For example, a portion of a prompt for predicting personalized health recommendations may be learned from labeled data with clinically validated personalized health recommendations. Soft prompting may comprise attaching fixed length vectors to the beginning of an embedding of an input before feeding the input into the model. The composition of a fixed length vector or vectors may be updated during a training operation such that the weights of the LLM are frozen during training and only the fixed length vector or vectors are updatable during training. In this manner, a prompt may be tuned to guide an LLM to resolve a subject input in an automated learning process rather than an iterative human prompt engineering process.
  • LLMs may provide a knowledge base that may be readily adapted to new information (e.g., a contextual corpus) or a specific task (e.g., a classification, regression, structured multi-modal output). LLMs may be trained on a large non-specific training sets in a generative pre-training operation. Generative pre-training may be used to establish a general understanding of a language such that a syntax, diction, vocabulary, and grammar of a language is learned by an LLM.
  • language may include natural language (e.g., vernacular English, vernacular Mandarin, vernacular Spanish) or domain specific language (e.g., jargon or connotation dictated by a field of use such as in a branch of science or knowledge context).
  • generative pre-training may comprise a series of training operations. For example, an LLM may be exposed to or trained on different domains over time.
  • a natural language corpus may be used to train an LLM prior to a domain specific language training.
  • generative pre-training may be performed for other modalities comprising image, video, tabular data, audio, or other data format with a learned or learnable embedding.
  • LLMs may be trained to perform a discriminate task such as regression (e.g., number prediction), classification (e.g., label prediction), or other specific output such as a semantic search, visual question answering, quality assurance, personalized health recommendations, or integrated wellness score calculation.
  • LLMs, or transformers may be used as feature extractors.
  • a complex (e.g., multi-modal, verbose, natural language) input may be fed into a trained transformer such that the output of trained transformer layer, which may be frozen (e.g., static model parameters), is used as input into a trainable head.
  • the trainable head is one or more deep learning layer (e.g., FCN, CNN) that may be trained on inputs into the trainable head (e.g., the extracted features) to predict a value of interest (e.g., personalized health recommendation).
  • Reinforcement Learning may be used to optimize models over time. Reinforcement learning may use human feedback to fine-tune models to output results in line with desired human or specific subject preference. Reinforcement learning from human feedback (RLHF) may comprise operations including supervised fine-tuning of an LLM, building a separate reward model, and optimizing the LLM with the reward model.
  • the supervised fine tuning may comprise training an LLM to provide outputs to prompts similar to those generated by a human.
  • the supervised fine-tuning may comprise collecting human demonstration data, for example, clinician health recommendations.
  • Building a separate reward model may comprise presenting a plurality of outputs of the fine-tuned LLM to a Attorney Docket No.67754-701.601 subject (e.g., clinician) for ranking.
  • the human ranking of the plurality of outputs may be used train the reward model to estimate a subject’s preference for an output of the fine-tuned LLM.
  • a subject may provide feedback (e.g., approval, rejection, rating) for a quality of a personalized health recommendation of virtual assistance.
  • Optimizing the LLM with the reward model may comprise a proximal policy optimization (PPO) to provide a loss to the LLM that may be used in updating a portion of the weights of the LLM.
  • PPO proximal policy optimization
  • RLHF may be used to gather preferences of a subject base (e.g., policyholders, clinicians) or a specific subject (e.g., single policyholder, clinician).
  • Federated Learning [0087]
  • federated learning (or collaborative learning) may be used to train one or more machine learning models. Federated learning may be used collaboratively train a machine learning model while maintaining decentralized data.
  • subject health data may be collected on a device (e.g., by an agent) and used to train or update a machine learning model on the device (e.g., personalized health recommendation model, virtual assistant).
  • Data used to train a machine learning model may be kept on a device and facilitate data privacy for a subject of the device.
  • Federated learning may aggregate learning data of devices (e.g., of a network of devices) and aggregate the learnings to update a centralized model.
  • the subject health data from a plurality of subjects may be used to train machine learning models associated with an agent of the plurality of subjects.
  • the updated centralized model may be distributed after training to devices implementing a version of the centralized model.
  • the centralized model may be a foundation or pre-trained model that is updated with learning data (e.g., learnings of agents of a network) and then optionally fine- tuned on subject data (e.g., subject data) upon deployment or re-deployment on subject devices.
  • learning data e.g., learnings of agents of a network
  • subject data e.g., subject data
  • federated learning may provide particular utility in protecting subject data while continuously improving the robustness of machine learning models subject to federated learning.
  • learning data of devices may be distributed among the devices directly rather than passed to a centralized model.
  • data shared for training e.g., one subject data to another subject device
  • the data shared for training does not reflect input features used for model inference and functions by aggregating locally computed (e.g., on the subject devices of a network) updates.
  • federated learning may implement iterative model averaging.
  • Centralized Learning may be used to train one or more machine learning models. Centralized learning may use local (e.g., one Attorney Docket No.67754-701.601 computing device) or distributed (e.g., a plurality of computing devices) data in training a machine learning model. For example, subject health data gathered in a network of agents as disclosed herein may be aggregated for provision to a centralized computing device.
  • any gathered subject health data may be de-identified, encrypted, or embedded to protect subject privacy.
  • Centralized learning may be used for pre-training or fine-tuning models for deployment to a computing device (optionally including the device used for centralized learning).
  • a machine learning model may be iteratively updated (e.g., re-trained) and deployed to devices implementing the machine learning model.
  • a large language model may be updated with subject health data, patient communications, or other medical knowledge to facilitate robustness.
  • a foundation model e.g., GPT, PaLM, LLaMa
  • the foundation model may be fine-tuned to accommodate specific tasks such as providing personalized health recommendations or acting as a virtual assistant (e.g., chatbot) to a subject while basing resolutions to the specific tasks on contemporary medical knowledge.
  • Life Insurance Product [0089] The techniques disclosed herein may be implemented in a pan-omic, AI driven, precision health service. This precision health service may be a life insurance product or life insurance policy. The life insurance product may pan-omic analysis across a spectrum of conditions such as cancer, cardiovascular, or neurological disease. The integrative approach disclosed herein may enable a policyholder of a life insurance product to receive targeted health insights or clinical recommendations. Such targeted health insights may provide particular utility in fostering proactive health management aimed at early prediction, disease prevention, or financial risk mitigation.
  • life insurance product may use AI or precision health data optimization that scrutinizes vast datasets enabling more precise diagnostics.
  • Personalized health recommendations or life insurance products as disclosed herein may appeal to any person seeking proactive health management. For example, families interested in hereditary risk management, older adults focused on proactive health management in retirement, or an individual with health concerns based on a known or perceived risk of imminent, nascent, eventual, or active changes in health status (e.g., ailing, ill, diseased, etc.).
  • a pan-omic analysis performed for a subject may become foundational to the knowledge of their agent as disclosed herein. The foundational knowledge may drive the personalized health recommendations provided to the subject.
  • pan-omic data may provide comprehensive insight into the health, predispositions, or well-being of a subject such Attorney Docket No.67754-701.601 that personalized health recommendations are highly unique or targeted.
  • the foundational knowledge may be supplemented with updated subject health data, lifestyle, or subject communications data (e.g., real-time natural language data, virtual assistant engagement). This may prepare the agent to provide personalized health recommendations in view of a diagnosis and prior subject behavior.
  • a life insurance product as disclosed herein may promote policy holder equity by providing personalized health recommendations. Such personalized health recommendations may be less prone to biases present in other treatment-focused or patient agnostic life insurance policies.
  • patient agnostic life insurance policies are those in which policies are based on populations (e.g., all policy holders), sub-populations (e.g., a subset of policy holders, demographics), or otherwise not unique to an individual policyholder.
  • a life insurance product as disclosed herein may provide access to precision health services. Precision health services may leverage advanced medical technologies (e.g., omic analysis) and personalized health recommendations to promote individualized health outcomes. For example, the personalized health recommendations may comprise preventive care plans, targeted therapies, or lifestyle interventions based on a subject’s pan-omic analysis and health goals.
  • the health goals may comprise policy holder selected health goals (e.g., desired goal weight, sleep quality, activity frequency) or policy holder suggested health goals (e.g., doctor suggestions, policy holder agent suggested).
  • a life insurance product as disclosed herein may promote longevity.
  • the life insurance product may comprise proactive (e.g., health monitoring, lifestyle interventions) and reactive (e.g., in view of health status change, diagnosis) components.
  • the pan-omic data leveraged herein may equip a subject with the knowledge (e.g., results of pan-omic analysis) and resources (e.g., personalized health recommendations) to facilitate informed decision making about health or lifestyle choices.
  • Personalized health recommendations for example, may ultimately extend lifespan and improve quality of life. Personalized health recommendations may mitigate potential health concerns or optimize well-being.
  • Proactive health measures may mitigate risks or severity of chronic health issues.
  • the mitigation of such risk may be enabled by pan-omic analysis indicating possible omic-based disease predispositions or nascent health problem (e.g., early indication of diabetes).
  • proactive health management may reduce costs for both policy holders and insurance providers (e.g., private, public). For example, early detection of medical issues, early warning of possible medical issues (e.g., predisposition), or personalized health recommendations (e.g., lifestyle changes, healthcare interventions) may reduce the risk of mitigatable medical events (e.g., severity reduction, avoidance of onset, decrease in recurrence risk).
  • an agent of networks are involved in simulating treatment outcomes and optimizing these outcomes through vetted requests by subjects interested in learning more from a bedrock of vendor products and services outside of the Network.
  • AI Agents assist subjects in vetting the best option to meet their request.
  • the network of agents disclosed herein may serve a similar function as a search engine. For example, a subject may provide a prompt expressing a health question or concern and receive personalized responses (e.g., precision health).
  • personalized health recommendations the network may facilitate directed advertisements or products for subjects.
  • a personalized health recommendation is an advertisement or product.
  • the subject may be made aware – by preference or by learned insight into the subject – of a product that may address a health concern of the subject.
  • a subject concerned by a family history of cancer may receive advertisements to a local cancer screening clinic.
  • a subject concerned with cardiovascular health may receive an advertisement for an exercise bike.
  • the product or recommendation is a health monitoring service (e.g., clinic, doctor, omic testing kit) or a health-related product (e.g., exercise equipment, supplement, personal diagnostic device).
  • a product or advertisement may be based on a subject health data, lifestyle, geographic location, or environmental needs.
  • An agent associated with the subject may be used to determine a quality, predicted efficacy, or relevance of an advertisement or product to subject health goals.
  • an advertisement or product may be a clinician presented to a subject.
  • the agent associated with the subject may suggest clinicians that are of high relevance to the subject.
  • the subject may provide feedback to a suggested product or advertisement. This feedback may be used to further align suggestions with subject preference.
  • machine learning e.g., RLHF
  • a life insurance product implementing the techniques disclosed herein may explicitly prohibit the use of omic data for underwriting purposes. This may promote subject privacy in view of ethical considerations.
  • a life insurance product as disclosed herein may comprise ancillary or voluntary benefits for group (e.g., employees, similar subjects) or individual.
  • ancillary or voluntary benefit may cover miscellaneous medical expenses such as ambulance transportation, blood, drugs, or medical supplies.
  • a blockchain (which may also be referred to e.g., as a distributed ledger or a shared ledger) is a technique that may be used for achieving a distributed consensus on validity or invalidity of information in a chain.
  • the blockchain provides a decentralized trust to participants and observers.
  • a blockchain is a distributed database, or ledger, in which a transactional record may be maintained at each node of a peer to peer network.
  • agents of a network as disclosed herein may produce simulations of case studies with tokens to test treatment strategies and efficacy, multiplying with every input.
  • Tokens may represent pieces of data that have value including in authentication or security processes. For example, tokens may be produced to grant an agent access to subject data without exposing credentials. Tokens may give an agent movement between human and machine without re-authenticating.
  • a token may represent a unit of value or a digital asset that may be issued or managed on a blockchain.
  • a token may represent rights that may be produced as reward incentives (wellness rewards) by the network to incentivize interaction and activity for pre-training.
  • the simulation of case studies may be favorable in the network due to the clean, validated, secure, and authentic data governance policy.
  • the distributed ledger may be comprised of groupings of transactions bundled together into a “block,” and ordered sequentially (thus the term “blockchain”). Nodes may join and leave the blockchain network over time and may obtain blocks that were propagated while the node was gone from peer nodes. Nodes may maintain addresses of other nodes and exchange addresses of known nodes with one another to facilitate the propagation of new information across the network in a decentralized, peer-to-peer manner. [0102] The nodes that share the ledger form what may be referred to as a distributed ledger network.
  • Consensus rules may include a mechanism to determine the order in which new blocks are added to the chain (e.g., through a proof-of-work system, proof-of- stake, etc.).
  • the agreed upon change is pushed out to each node so that each node maintains a copy (e.g., identical copy) of the updated distributed ledger.
  • additions to the blockchain that satisfy the consensus rules may be propagated from nodes that have validated the addition to other nodes that the validating node is aware of.
  • the distributed ledger reflects the new change as stored on all nodes, and it may be said that distributed consensus has been reached with respect to the new block and the information contained therein. Any change that does not satisfy the consensus rule may be disregarded by validating nodes that receive the change and may not be propagated to other nodes. Accordingly, unlike a central authority, a single party cannot unilaterally alter the distributed ledger, unless the single party can do so in a way that satisfies the consensus rules. This inability to modify past transactions leads to blockchains being generally described as trusted, secure, and immutable. [0104]
  • the validation activities of nodes applying consensus rules on a blockchain network may take various forms.
  • a smart contract may be a computer protocol that enables the automatic execution or enforcement of an agreement between different parties.
  • the smart contract may be computer code that is located at a particular address on the blockchain.
  • the smart contract may run automatically in response to a participant in the blockchain sending funds (e.g., a cryptocurrency such as bitcoin, ether, or other digital or virtual currencies) to an address where the smart contract is stored.
  • funds e.g., a cryptocurrency such as bitcoin, ether, or other digital or virtual currencies
  • the smart contract may include one or more trigger conditions, that, when satisfied, correspond to one or more actions. For some smart contracts, which actions from the one or more actions are performed may be determined based at least in part on one or more decision conditions. In some cases, data streams may be routed to the smart Attorney Docket No.67754-701.601 contract so that the smart contract may detect that a trigger condition has occurred or analyze a decision condition.
  • Blockchains may be deployed as private (e.g., permissioned ledgers) blockchains that keep chain data private among a group of entities authorized to participate in the blockchain network.
  • each transaction within a block may be assigned a hash value (e.g., an output of a cryptographic hash function, such as SHA-256 or MD5).
  • a hash value e.g., an output of a cryptographic hash function, such as SHA-256 or MD5
  • hash values may then be combined together utilizing data storage and cryptographic techniques (e.g., a Merkle Tree) to generate a hash value representative of the entire new block, and consequently the transactions stored in the block.
  • This hash value may then be combined with the hash value of the previous block to form a hash value included in the header of the new block, thereby cryptographically linking the new block to the blockchain.
  • the precise value utilized in the header of the new block may be dependent on the hash value for each transaction in the new block, as well as the hash value for each transaction in every prior block.
  • information stored in blockchains may be trusted (e.g., at least partially trusted), because the hash value generated for the new block and a nonce value (an arbitrary number used once) may be used as inputs into a cryptographic puzzle.
  • the cryptographic puzzle may have a difficulty set by the nodes connected to the blockchain network, or the difficulty may be set by administrators of the blockchain network.
  • a solving node uses the hash value generated for the new block and repeatedly changes the value of the nonce until a solution for the puzzle is found. For example, finding the solution to the cryptographic puzzle may involve finding the nonce value that meets certain criteria (e.g., the nonce value begins with five zeros). [0109] When a solution to the cryptographic puzzle is found, the solving node publishes the solution and the other nodes may then verify that the solution is valid. Since the solution may depend on the particular hash values for each transaction within the blockchain, if the solving node attempts to modify any transaction stored in the blockchain, the solution may not be verified by the other nodes.
  • ML machine learning
  • RL reinforcement learning
  • machine learning machine learning techniques
  • machine learning operation machine learning model
  • ML may generally involve identifying and recognizing patterns in existing data in order to facilitate making predictions for subsequent data.
  • ML may include a ML model (which may include, for example, a ML algorithm).
  • Machine learning whether analytical or statistical in nature, may provide deductive or abductive inference based on real or simulated data.
  • the ML model may be a trained model.
  • ML techniques may comprise one or more supervised, semi-supervised, self-supervised, or unsupervised ML techniques.
  • an ML model may be a trained model that is trained through supervised learning (e.g., various parameters are determined as weights or scaling factors).
  • ML may comprise one or more of regression analysis, regularization, classification, dimensionality reduction, ensemble learning, meta learning, association rule learning, cluster analysis, anomaly detection, deep learning, or ultra-deep learning.
  • ML may comprise: k-means, k-means clustering, k-nearest neighbors, learning vector quantization, linear regression, non-linear regression, least squares regression, partial least squares regression, logistic regression, stepwise regression, multivariate adaptive regression splines, ridge regression, principal component regression, least absolute shrinkage and selection operation (LASSO), least angle regression, canonical correlation analysis, factor analysis, independent component analysis, linear discriminant analysis, multidimensional scaling, non-negative matrix factorization, principal components analysis, principal coordinates analysis, projection pursuit, Sammon mapping, t- distributed stochastic neighbor embedding, AdaBoosting, boosting, gradient boosting, bootstrap aggregation, ensemble averaging, decision trees, conditional decision trees, boosted decision trees, gradient boosted decision trees, random forests, stacked generalization, Bayesian networks, Bayesian belief networks, na ⁇ ve Bayes, Gaussian na ⁇ ve Bayes, multinomial na ⁇ ve Bayes, hidden Markov models, hierarchical hidden
  • Training the ML model may include, in some cases, selecting one or more untrained data models to train using a training data set.
  • the selected untrained data models may include any type of untrained ML models for supervised, semi-supervised, self-supervised, or unsupervised machine learning.
  • the selected untrained data models may be specified based upon input (e.g., subject input) specifying relevant parameters to use as predicted variables or other variables to use as potential explanatory variables.
  • the selected untrained data models may be specified to generate an output (e.g., a prediction) based upon the input.
  • Conditions for training the ML model from the selected untrained data models may likewise be selected, such as limits on the ML model complexity or limits on the ML model refinement past a certain point.
  • the ML model may be trained (e.g., via a computer system such as a server) using the training data set.
  • a first subset of the training data set may be selected to train the ML model.
  • the selected untrained data models may then be trained on the first subset of training data set using appropriate ML techniques, based upon the type of ML model selected and any conditions specified for training the ML model.
  • the selected untrained data models may be trained using additional computing resources (e.g., cloud computing resources). Such training may continue, in some cases, until at least one aspect of the ML model is validated and meets selection criteria to be used as a predictive model.
  • one or more aspects of the ML model may be validated using a second subset of the training data set (e.g., distinct from the first subset of the training data set) to determine accuracy and robustness of the ML model.
  • Such validation may include applying the ML model to the second subset of the training data set to make predictions derived from the second subset of the training data.
  • the ML model may then be evaluated to determine whether performance is sufficient based upon the derived predictions.
  • the sufficiency criteria applied to the ML model may vary depending upon the size of the training data set available for training, the performance of previous iterations of trained models, or subject-specified performance requirements. If the ML model does not achieve sufficient performance, additional training may be performed.
  • Additional training may include refinement of the ML model or retraining on a different first subset of the training dataset, after which the new ML model may again be validated and assessed.
  • the ML may be stored for present or future use.
  • the ML model may be stored as sets of parameter values or weights for analysis of further input (e.g., further relevant parameters to use as further predicted Attorney Docket No.67754-701.601 variables, further explanatory variables, further subject interaction data, etc.), which may also include analysis logic or indications of model validity in some instances.
  • a plurality of ML models may be stored for generating predictions under different sets of input data conditions.
  • the network 230 of the computer system 200 may be used to interaction with the network 130 of agents as disclosed herein. Additionally, or alternatively the computer system 200 may implement training of machine learning models (e.g., federated learning, centralized learning) using the one or more processors 201 and data stored in the one or more storage devices 235. In another example, a subject (e.g., subject 110) or clinician (e.g., clinician 150) may interact with their agent (e.g., virtual assistant, nurse navigator) through the display 232.
  • Computer system 200 may include one or more processors 201, a memory 203, and a storage 208 that communicate with each other, and with other components, via a bus 240.
  • the bus 240 may also link a display 232, one or more input devices 233 (which may, for example, include a keypad, a keyboard, a mouse, a stylus, etc.), one or more output devices 234, one or more storage devices 235, and various tangible storage media 236. All of these elements may interface directly or via one or more interfaces or adaptors to the bus 240. For instance, the various tangible storage media 236 can interface with the bus 240 via storage medium interface 226.
  • Computer system 200 may have any suitable physical form, including but not limited to one or more integrated circuits (ICs), printed circuit boards (PCBs), mobile handheld devices (such as mobile telephones or PDAs), laptop or notebook computers, distributed computer systems, computing grids, or servers.
  • ICs integrated circuits
  • PCBs printed circuit boards
  • mobile handheld devices such as mobile telephones or PDAs
  • laptop or notebook computers distributed computer systems, computing grids, or servers.
  • Computer system 200 may provide functionality for the components depicted in FIG.2 as a result of the processor(s) 201 executing non-transitory, processor-executable instructions embodied in one or more tangible computer-readable storage media, such as memory 203, storage 208, storage devices 235, or storage medium 236.
  • the computer-readable media may store software that implements particular embodiments, and processor(s) 201 may execute the software.
  • Memory 203 may read the software from one or more other computer-readable media (such as mass storage device(s) 235, 236) or from one or more other sources through a suitable interface, such as network interface 220.
  • the software may cause processor(s) 201 to carry out one or more processes or one or more steps of one or more processes described or illustrated herein.
  • the memory 203 may include various components (e.g., machine readable media) including, but not limited to, a random access memory component (e.g., RAM 204) (e.g., static RAM (SRAM), dynamic RAM (DRAM), ferroelectric random access memory (FRAM), phase- change random access memory (PRAM), etc.), a read-only memory component (e.g., ROM 205), and any combinations thereof.
  • ROM 205 may act to communicate data and instructions unidirectionally to processor(s) 201
  • RAM 204 may act to communicate data and instructions bidirectionally with processor(s) 201.
  • ROM 205 and RAM 204 may include any suitable tangible computer-readable media described below.
  • a basic input/output system 206 (BIOS), including basic routines that help to transfer information between elements within computer system 200, such as during start-up, may be stored in the memory 203.
  • Fixed storage 208 is connected bidirectionally to processor(s) 201, optionally through storage control unit 207. Fixed storage 208 provides additional data storage capacity and may also include any suitable tangible computer-readable media described herein. Storage 208 may be used to store operating system 209, executable(s) 210, data 211, applications 212 (application programs), and the like.
  • Storage 208 can also include an optical disk drive, a solid-state memory device (e.g., flash-based systems), or a combination of any of the above. Information in storage 208 may, in appropriate cases, be incorporated as virtual memory in memory 203.
  • storage device(s) 235 may be removably interfaced with computer system 200 (e.g., via an external port connector (not shown)) via a storage device interface 225.
  • storage device(s) 235 and an associated machine-readable medium may provide non- volatile or volatile storage of machine-readable instructions, data structures, program modules, or other data for the computer system 200.
  • Computer system 200 may also include an input device 233.
  • a user e.g., subject 110, clinician 150 of computer system 200 may enter commands or other information into computer system 200 via input device(s) 233.
  • Examples of an input device(s) 233 include, but are not limited to, an alpha-numeric input device (e.g., a keyboard), a pointing device (e.g., a mouse or touchpad), a touchpad, a touch screen, a multi-touch screen, a joystick, a stylus, a gamepad, an audio input device (e.g., a microphone, a voice response system, etc.), an optical scanner, a video or still image capture device (e.g., a camera), and any combinations thereof.
  • the input device is a Kinect, Leap Motion, or the like.
  • Examples of a network 230 or network segment 230 include, but are not limited to, a distributed computing system, a cloud computing system, a wide area network (WAN) (e.g., the Internet, an enterprise network), a local area network (LAN) (e.g., a network associated with an office, a building, a campus or other relatively small geographic space), a telephone network, a direct connection between two computing devices, a peer-to-peer network, and any combinations thereof.
  • a network, such as network 230 may employ a wired or a wireless mode of communication. In general, any network topology may be used.
  • Information and data can be displayed through a display 232.
  • Examples of a display 232 include, but are not limited to, a cathode ray tube (CRT), a liquid crystal display (LCD), a thin film transistor liquid crystal display (TFT-LCD), an organic liquid crystal display (OLED) such as a passive-matrix OLED (PMOLED) or active-matrix OLED (AMOLED) display, a plasma display, and any combinations thereof.
  • the display 232 can interface to the processor(s) 201, memory 203, and fixed storage 208, as well as other devices, such as input device(s) 233, via the bus 240.
  • the display 232 is linked to the bus 240 via a video interface 222, and transport of data between the display 232 and the bus 240 can be controlled via the graphics control 221.
  • the display is a video projector.
  • the display is a head-mounted display (HMD) such as a VR headset.
  • HMD head-mounted display
  • suitable VR headsets include, by way of non-limiting examples, HTC Vive, Oculus Rift, Samsung Gear VR, Microsoft HoloLens, Razer OSVR, FOVE VR, Zeiss VR One, Avegant Glyph, Freefly VR headset, and the like.
  • the display is a combination of devices such as those disclosed herein.
  • computer system 200 may include one or more other peripheral output devices 234 including, but not limited to, an audio speaker, a printer, a storage device, and any combinations thereof.
  • reference to a computer-readable medium Attorney Docket No.67754-701.601 may encompass a circuit (such as an IC) storing software for execution, a circuit embodying logic for execution, or both, where appropriate.
  • the present disclosure encompasses any suitable combination of hardware, software, or both.
  • Those of skill in the art will appreciate that the various illustrative logical blocks, modules, circuits, and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both. To clearly illustrate this interchangeability of hardware and software, various illustrative components, blocks, modules, circuits, and steps have been described above generally in terms of their functionality.
  • a processor may also be implemented as a combination of computing devices, e.g., a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration.
  • the steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by one or more processor(s), or in a combination of the two.
  • a software module may reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.
  • An exemplary storage medium is coupled to the processor such the processor can read information from, and write information to, the storage medium.
  • the storage medium may be integral to the processor.
  • the processor and the storage medium may reside in an ASIC.
  • the ASIC may reside in a user terminal.
  • the processor and the storage medium may reside as discrete components in a user terminal.
  • suitable computing devices include, by way of non-limiting examples, server computers, desktop computers, laptop computers, notebook computers, sub-notebook computers, netbook computers, netpad computers, set-top computers, media streaming devices, handheld computers, Internet appliances, mobile smartphones, tablet computers, personal digital assistants, video game consoles, and vehicles.
  • Computing devices may further comprise televisions, video players, and digital music players with optional computer Attorney Docket No.67754-701.601 network connectivity are suitable for use in the system described herein.
  • Suitable tablet computers include those with booklet, slate, and convertible configurations, known to those of skill in the art.
  • the computing device includes an operating system configured to perform executable instructions.
  • the operating system is, for example, software, including programs and data, which manages the device’s hardware and provides services for execution of applications.
  • a server operating system may include, by way of non-limiting examples, FreeBSD, OpenBSD, NetBSD ® , Linux, Apple ® Mac OS X Server ® , Oracle ® Solaris ® , Windows Server ® , and Novell ® NetWare ® .
  • a suitable personal computer operating system may include, by way of non- limiting examples, Microsoft ® Windows ® , Apple ® Mac OS X ® , UNIX ® , and UNIX-like operating systems such as GNU/Linux ® .
  • the operating system is provided by cloud computing.
  • a suitable mobile smartphone operating system include, by way of non-limiting examples, Nokia ® Symbian ® OS, Apple ® iOS ® , Research In Motion ® BlackBerry OS ® , Google ® Android ® , Microsoft ® Windows Phone ® OS, Microsoft ® Windows Mobile ® OS, Linux ® , and Palm ® WebOS ® .
  • Suitable media streaming device operating systems may include, by way of non- limiting examples, Apple TV ® , Roku ® , Boxee ® , Google TV ® , Google Chromecast ® , Amazon Fire ® , and Samsung ® HomeSync ® .
  • Suitable video game console operating system may include, by way of non-limiting examples, Sony ® PS3 ® , Sony ® PS4 ® , Sony ® PS5 ® , Microsoft ® Xbox 360 ® , Microsoft ® Xbox One, Microsoft ® Xbox Series X, Microsoft ® Xbox Series S, Nintendo ® Wii ® , Nintendo ® Wii U ® , Nintendo ® Switch TM , and Ouya ® .
  • Another aspect of the disclosure herein describes a non-transitory, computer-readable medium comprising executable instructions, wherein when a processor, when executing the executable instructions, performs a method as described herein.
  • Web application [0134]
  • a computer program includes a web application.
  • a web application in various embodiments, utilizes one or more software frameworks and one or more database systems.
  • a web application is created upon a software framework such as Microsoft ® .NET or Ruby on Rails (RoR).
  • a web application utilizes one or more database systems including, by way of non- limiting examples, relational, non-relational, object oriented, associative, XML, and document oriented database systems.
  • suitable relational database systems include, by way of non-limiting examples, Microsoft ® SQL Server, mySQLTM, and Oracle ® .
  • a web application in various embodiments, is written in one or more versions of one or more languages.
  • a web application may be written in one or more markup languages, presentation Attorney Docket No.67754-701.601 definition languages, client-side scripting languages, server-side coding languages, database query languages, or combinations thereof.
  • a web application is written to some extent in a markup language such as Hypertext Markup Language (HTML), Extensible Hypertext Markup Language (XHTML), or eXtensible Markup Language (XML).
  • a web application is written to some extent in a presentation definition language such as Cascading Style Sheets (CSS).
  • a web application is written to some extent in a client- side scripting language such as Asynchronous JavaScript and XML (AJAX), Flash ® ActionScript, JavaScript, or Silverlight ® .
  • AJAX Asynchronous JavaScript and XML
  • Flash ® ActionScript JavaScript
  • Silverlight ® a web application is written to some extent in a server-side coding language such as Active Server Pages (ASP), ColdFusion ® , Perl, JavaTM, JavaServer Pages (JSP), Hypertext Preprocessor (PHP), PythonTM, Ruby, Tcl, Smalltalk, WebDNA ® , or Groovy.
  • a web application is written to some extent in a database query language such as Structured Query Language (SQL).
  • SQL Structured Query Language
  • a web application integrates enterprise server products such as IBM ® Lotus Domino ® .
  • a web application includes a media player element.
  • a media player element utilizes one or more of many suitable multimedia technologies including, by way of non-limiting examples, Adobe ® Flash ® , HTML 5, Apple ® QuickTime ® , Microsoft ® Silverlight ® , JavaTM, and Unity ® .
  • an application provision system comprises one or more databases 300 accessed by a relational database management system (RDBMS) 310.
  • RDBMS relational database management system
  • Suitable RDBMSs include Firebird, MySQL, PostgreSQL, SQLite, Oracle Database, Microsoft SQL Server, IBM DB2, IBM Informix, SAP Sybase, Teradata, and the like.
  • the application provision system further comprises one or more application severs 320 (such as Java servers, .NET servers, PHP servers, and the like) and one or more web servers 330 (such as Apache, IIS, GWS and the like).
  • the web server(s) optionally expose one or more web services via app application programming interfaces (APIs) 340.
  • APIs app application programming interfaces
  • a computer program includes a standalone application, which is a program that is run as an independent computer process, not an add-on to an existing process, e.g., not a plug-in.
  • a standalone applications may be compiled.
  • a compiler is a computer program(s) that transforms source code written in a programming language into binary object code such as assembly language or machine code.
  • Suitable compiled programming languages include, by way of non-limiting examples, C, C++, Objective-C, COBOL, Delphi, Eiffel, JavaTM, Lisp, PythonTM, Visual Basic, and VB .NET, or combinations thereof. Compilation is often performed, at least in part, to create an executable program.
  • a computer program includes one or more executable complied applications.
  • Web browser plug-in [0142]
  • the computer program includes a web browser plug-in (e.g., extension, etc.).
  • a plug-in is one or more software components that add specific functionality to a larger software application.
  • Web browsers are software applications, designed for use with network-connected computing devices, for retrieving, presenting, and traversing information resources on the World Wide Web. Suitable web browsers include, by way of non-limiting examples, Microsoft ® Internet Explorer ® , Mozilla ® Firefox ® , Google ® Chrome, Apple ® Safari ® , Opera Software ® Opera ® , and KDE Konqueror. In some embodiments, the web browser is a mobile web browser.
  • Mobile web browsers are designed for use on mobile computing devices including, by way of non-limiting examples, handheld computers, tablet computers, netbook computers, subnotebook computers, smartphones, music players, personal digital assistants (PDAs), and handheld video game systems.
  • Suitable mobile web browsers include, by way of non-limiting examples, Google ® Android ® browser, RIM BlackBerry ® Browser, Apple ® Safari ® , Palm ® Blazer, Palm ® WebOS ® Browser, Mozilla ® Firefox ® for mobile, Microsoft ® Internet Explorer ® Mobile, Amazon ® Kindle ® Basic Web, Nokia ® Browser, Opera Software ® Opera ® Mobile, and Sony ® PSPTM browser.
  • the platforms, systems, media, and methods disclosed herein include software, server, or database modules, or use of the same.
  • software modules may be created using machines, software, and languages.
  • the software modules disclosed herein are implemented in a multitude of ways.
  • a software module comprises a file, a section of code, a programming object, a programming structure, a distributed computing resource, a cloud computing resource, or combinations thereof.
  • a software module comprises a plurality of files, a plurality of sections of code, a plurality of programming objects, a plurality of programming structures, a plurality of distributed computing resources, a plurality of cloud computing resources, or Attorney Docket No.67754-701.601 combinations thereof.
  • the one or more software modules comprise, by way of non-limiting examples, a web application, a mobile application, a standalone application, and a distributed or cloud computing application.
  • software modules are in one computer program or application. In other embodiments, software modules are in more than one computer program or application. In some embodiments, software modules are hosted on one machine. In other embodiments, software modules are hosted on more than one machine.
  • software modules are hosted on a distributed computing platform such as a cloud computing platform.
  • software modules are hosted on one or more machines in one location.
  • software modules are hosted on one or more machines in more than one location.
  • Databases [0146]
  • the platforms, systems, media, and methods disclosed herein include one or more databases, or use of the same.
  • databases are suitable for storage and retrieval of health data, model parameters, or any combination thereof.
  • suitable databases include, by way of non-limiting examples, relational databases, non-relational databases, object oriented databases, object databases, entity-relationship model databases, associative databases, XML databases, document oriented databases, and graph databases.
  • a database is Internet-based.
  • a database is web-based.
  • a database is cloud computing based.
  • a database is a distributed database.
  • a database is based on one or more local computer storage devices. Data transmission [0147] The subject matter described herein, including methods and systems as described herein and may be configured to be performed in one or more facilities at one or more locations. Facility locations are not limited by country and include any country or territory. In some instances, one or more steps are performed in a different country than another step of the method.
  • one or more method steps involving a computer system are performed in a different country than another step of the methods provided herein.
  • data processing and storage are performed in a different country or location than one or more steps of the methods described herein.
  • one or more products or data are transferred from one or more of the facilities to one or more different facilities for analysis or further analysis.
  • Data includes, but is not limited to, information regarding the stratification of a subject, and any data produced by the methods disclosed herein.
  • the methods and systems Attorney Docket No.67754-701.601 described herein, the subject information is compiled, and a subsequent data transmission step will transmit or store the subject information.
  • machine learning As used in this specification and the appended claims, the terms “machine learning,” “machine learning techniques”, “machine learning algorithm,” “machine learning operation,” and “machine learning model” generally refer to any system or analytical or statistical procedure that may progressively improve computer performance of a task.
  • Example 1 Life Insurance Policy
  • FIG.5 shows how each step of a policyholder’s journey from initial inquiry to precision health services is clearly outlined in comparison to a current life insurance policy.
  • the comparison Attorney Docket No.67754-701.601 demonstrates how personalized omics, lifestyle, and environmental factors can play a critical role in how life policyholders can proactively manage their wellbeing for a healthier lifespan.
  • Example 2 Personalized Health Recommendation Input
  • GRS Genetic Risk Score
  • LRI Lifestyle Risk Index
  • IWS Integrated Wellness Score
  • Operation 3 Health Data Score (HDS) [0162] Given the health conditions mentioned: short telomeres: indicator of accelerated biological aging and increased disease risk. For simplicity, let's score this as 8, indicating higher risk, high blood pressure and high cholesterol: already factored into the LRI but also contribute to an elevated HDS – Assign an HDS of 8, reflecting significant concern.
  • HDS Health Data Score
  • the health data is uploaded to the network using blockchain technology.
  • a chatbot of the agent trained in omic data and patient communication engages with the subject.
  • the chatbot engages with the subject regarding their questions or concerns spawning from their pan-omic analysis. All communication is recorded and saved as an input and uploaded into the agent for record-keeping and pre-training. The test results and significance are explained, and potential health goals and strategies are also tailored to the individual’s genetic makeup. The discussion lasts about 45 minutes with Q&A.
  • the subject’s agent is inducted into a secure, agent-only social media network for the purpose of treatment optimization. This social media network functions as an autonomous interactive device where agents simulate the collaborative dynamics of social media networks.
  • the Pre-Training AI Agent Social Media Network focuses on the exchange of pan-omic insights, real-time de-centralized data (from clinicians and subjects), and vetted products, services, and solutions; ultimately pre-training agents in preparation for precision medicine and precision health treatments should a diagnosis occur.
  • agents engage in pre-training, exchanging real-time data and medical insights.
  • This pre-training process ensures that agents can provide subjects and their physician (clinicians) with optimized treatment strategies, vetted products and solutions, especially in the event of a diagnosis and working in conjunction with the post-diagnosis stage outlined in the insurance policy (cancer, cardio, neuro, etc.).
  • the Pre-Training agent social network facilitates ongoing learning for agents of the network, ensuring they stay updated with the latest AI generative technology, medical advancements and treatment protocols, leading to continuous improvement.
  • their agent can immediately utilize its pre-training to engage appropriate post-diagnostic testing protocols and synthesized data for reporting to clinicians of potential treatments and to the post-diagnostic labs conducting pan-omics testing, thereby minimizing the time to intervention and improving health outcomes.

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Public Health (AREA)
  • Data Mining & Analysis (AREA)
  • General Health & Medical Sciences (AREA)
  • Epidemiology (AREA)
  • Biomedical Technology (AREA)
  • Databases & Information Systems (AREA)
  • Primary Health Care (AREA)
  • Pathology (AREA)
  • Physics & Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Biotechnology (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Biophysics (AREA)
  • Theoretical Computer Science (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Evolutionary Biology (AREA)
  • Proteomics, Peptides & Aminoacids (AREA)
  • Chemical & Material Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Molecular Biology (AREA)
  • Genetics & Genomics (AREA)
  • Artificial Intelligence (AREA)
  • Bioethics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Computation (AREA)
  • Software Systems (AREA)
  • Medical Treatment And Welfare Office Work (AREA)

Abstract

L'invention divulgue des systèmes et des procédés de fourniture de recommandations de santé personnalisées. Des recommandations de santé personnalisées peuvent être fournies par un algorithme d'apprentissage automatique implémenté dans un agent associé à un sujet. L'agent associé au sujet peut être dans un réseau d'agents qui peuvent collaborer pour fournir des recommandations de santé personnalisées améliorées pour des sujets du réseau.
PCT/US2025/023067 2024-04-05 2025-04-03 Gestion de santé panomique intégrative Pending WO2025212971A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US202463575551P 2024-04-05 2024-04-05
US63/575,551 2024-04-05

Publications (1)

Publication Number Publication Date
WO2025212971A1 true WO2025212971A1 (fr) 2025-10-09

Family

ID=97268012

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2025/023067 Pending WO2025212971A1 (fr) 2024-04-05 2025-04-03 Gestion de santé panomique intégrative

Country Status (1)

Country Link
WO (1) WO2025212971A1 (fr)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160019336A1 (en) * 2006-05-03 2016-01-21 Population Diagnostics, Inc. Evaluating genetic disorders
WO2020023959A1 (fr) * 2018-07-27 2020-01-30 University Of Miami Système et procédé pour déterminations de trouble de l'œil à base d'ia
US20200082945A1 (en) * 2016-12-22 2020-03-12 Basehealth, Inc. Scoring and mitigating health risks
US20200411199A1 (en) * 2018-01-22 2020-12-31 Cancer Commons Platforms for conducting virtual trials
KR20220051266A (ko) * 2019-09-06 2022-04-26 스포츠 데이타 랩스, 인코포레이티드 시뮬레이션된 동물 데이터 및 모델을 생성하기 위한 시스템
US20230054253A1 (en) * 2021-08-06 2023-02-23 Food Rx and AI, Inc. Methods and systems for multi-omic interventions
WO2024059341A2 (fr) * 2022-09-16 2024-03-21 Progentec Diagnostics, Inc. Système et procédé d'intelligence artificielle pour l'évaluation du pronostic des troubles auto-immuns

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160019336A1 (en) * 2006-05-03 2016-01-21 Population Diagnostics, Inc. Evaluating genetic disorders
US20200082945A1 (en) * 2016-12-22 2020-03-12 Basehealth, Inc. Scoring and mitigating health risks
US20200411199A1 (en) * 2018-01-22 2020-12-31 Cancer Commons Platforms for conducting virtual trials
WO2020023959A1 (fr) * 2018-07-27 2020-01-30 University Of Miami Système et procédé pour déterminations de trouble de l'œil à base d'ia
KR20220051266A (ko) * 2019-09-06 2022-04-26 스포츠 데이타 랩스, 인코포레이티드 시뮬레이션된 동물 데이터 및 모델을 생성하기 위한 시스템
US20230054253A1 (en) * 2021-08-06 2023-02-23 Food Rx and AI, Inc. Methods and systems for multi-omic interventions
WO2024059341A2 (fr) * 2022-09-16 2024-03-21 Progentec Diagnostics, Inc. Système et procédé d'intelligence artificielle pour l'évaluation du pronostic des troubles auto-immuns

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
GODARD BÉATRICE, RAEBURN SANDY, PEMBREY MARCUS, BOBROW MARTIN, FARNDON PETER, AYMÉ SÉGOLÈNE: "Genetic information and testing in insurance and employment: technical, social and ethical issues", EUROPEAN JOURNAL OF HUMAN GENETICS, KARGER, BASEL, CH, vol. 11, no. S2, 1 December 2003 (2003-12-01), CH , pages S123 - S142, XP093364973, ISSN: 1018-4813, DOI: 10.1038/sj.ejhg.5201117 *

Similar Documents

Publication Publication Date Title
Getzen et al. Mining for equitable health: Assessing the impact of missing data in electronic health records
US20240186023A1 (en) Platforms for conducting virtual trials
Li et al. Open challenges and opportunities in federated foundation models towards biomedical healthcare
Sharon The Googlization of health research: from disruptive innovation to disruptive ethics
Coiera et al. Evidence synthesis, digital scribes, and translational challenges for artificial intelligence in healthcare
Chawla et al. Bringing big data to personalized healthcare: a patient-centered framework
Vollmer et al. Machine learning and AI research for patient benefit: 20 critical questions on transparency, replicability, ethics and effectiveness
Han et al. Evaluation of GPT-4 for 10-year cardiovascular risk prediction: insights from the UK Biobank and KoGES data
Klang et al. A strategy for cost-effective large language model use at health system-scale
Ng et al. Personalized treatment options for chronic diseases using precision cohort analytics
US20230260665A1 (en) Modeling for complex outcomes using similarity and machine learning algorithms
Paquette et al. Machine learning support for decision-making in kidney transplantation: step-by-step development of a technological solution
US11328825B1 (en) Machine learning techniques for identifying opportunity patients
US12475995B2 (en) Decision support systems for determining conformity with medical care quality standards
Singhal et al. Opportunities and challenges for biomarker discovery using electronic health record data
Lenatti et al. A novel method to derive personalized minimum viable recommendations for type 2 diabetes prevention based on counterfactual explanations
Mundt et al. AI‐driven personalized nutrition: integrating omics, ethics, and digital health
Bergquist et al. Piloting a model-to-data approach to enable predictive analytics in health care through patient mortality prediction
Scott et al. Informatics for Health 2017: Advancing both science and practice
Chava Revolutionizing Healthcare Systems with Next-Generation Technologies: The Role of Artificial Intelligence, Cloud Infrastructure, and Big Data in Driving Patient-Centric Innovation
Razmi AI Doctor: The Rise of Artificial Intelligence in Healthcare-A Guide for Users, Buyers, Builders, and Investors
Zong et al. Advancing efficacy prediction for electronic health records based emulated trials in repurposing heart failure therapies
Layton A heart-to-heart with ChatGPT: AI applications in hypertension
Panesar Precision Health and Artificial Intelligence
US20240274290A1 (en) Predicting changes in risk based on interventions

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 25783357

Country of ref document: EP

Kind code of ref document: A1