WO2023249372A1 - 딥러닝 모델의 학습을 위한 의료 데이터의 처리 방법, 프로그램 및 장치 - Google Patents
딥러닝 모델의 학습을 위한 의료 데이터의 처리 방법, 프로그램 및 장치 Download PDFInfo
- Publication number
- WO2023249372A1 WO2023249372A1 PCT/KR2023/008528 KR2023008528W WO2023249372A1 WO 2023249372 A1 WO2023249372 A1 WO 2023249372A1 KR 2023008528 W KR2023008528 W KR 2023008528W WO 2023249372 A1 WO2023249372 A1 WO 2023249372A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- data
- preliminary
- space
- label
- generate
- 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.)
- Ceased
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/05—Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves
- A61B5/055—Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves involving electronic [EMR] or nuclear [NMR] magnetic resonance, e.g. magnetic resonance imaging
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R33/00—Arrangements or instruments for measuring magnetic variables
- G01R33/20—Arrangements or instruments for measuring magnetic variables involving magnetic resonance
- G01R33/44—Arrangements or instruments for measuring magnetic variables involving magnetic resonance using nuclear magnetic resonance [NMR]
- G01R33/48—NMR imaging systems
- G01R33/54—Signal processing systems, e.g. using pulse sequences ; Generation or control of pulse sequences; Operator console
- G01R33/56—Image enhancement or correction, e.g. subtraction or averaging techniques, e.g. improvement of signal-to-noise ratio and resolution
- G01R33/5608—Data processing and visualization specially adapted for MR, e.g. for feature analysis and pattern recognition on the basis of measured MR data, segmentation of measured MR data, edge contour detection on the basis of measured MR data, for enhancing measured MR data in terms of signal-to-noise ratio by means of noise filtering or apodization, for enhancing measured MR data in terms of resolution by means for deblurring, windowing, zero filling, or generation of gray-scaled images, colour-coded images or images displaying vectors instead of pixels
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
- G06V10/765—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects using rules for classification or partitioning the feature space
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/88—Image or video recognition using optical means, e.g. reference filters, holographic masks, frequency domain filters or spatial domain filters
- G06V10/89—Image or video recognition using optical means, e.g. reference filters, holographic masks, frequency domain filters or spatial domain filters using frequency domain filters, e.g. Fourier masks implemented on spatial light modulators
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/70—Labelling scene content, e.g. deriving syntactic or semantic representations
-
- 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
- G16H30/00—ICT specially adapted for the handling or processing of medical images
- G16H30/20—ICT specially adapted for the handling or processing of medical images for handling medical images, e.g. DICOM, HL7 or PACS
-
- 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
- G16H30/00—ICT specially adapted for the handling or processing of medical images
- G16H30/40—ICT specially adapted for the handling or processing of medical images for processing medical images, e.g. editing
-
- 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
- G16H50/70—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
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R33/00—Arrangements or instruments for measuring magnetic variables
- G01R33/20—Arrangements or instruments for measuring magnetic variables involving magnetic resonance
- G01R33/44—Arrangements or instruments for measuring magnetic variables involving magnetic resonance using nuclear magnetic resonance [NMR]
- G01R33/48—NMR imaging systems
- G01R33/54—Signal processing systems, e.g. using pulse sequences ; Generation or control of pulse sequences; Operator console
- G01R33/56—Image enhancement or correction, e.g. subtraction or averaging techniques, e.g. improvement of signal-to-noise ratio and resolution
- G01R33/561—Image enhancement or correction, e.g. subtraction or averaging techniques, e.g. improvement of signal-to-noise ratio and resolution by reduction of the scanning time, i.e. fast acquiring systems, e.g. using echo-planar pulse sequences
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R33/00—Arrangements or instruments for measuring magnetic variables
- G01R33/20—Arrangements or instruments for measuring magnetic variables involving magnetic resonance
- G01R33/44—Arrangements or instruments for measuring magnetic variables involving magnetic resonance using nuclear magnetic resonance [NMR]
- G01R33/48—NMR imaging systems
- G01R33/54—Signal processing systems, e.g. using pulse sequences ; Generation or control of pulse sequences; Operator console
- G01R33/56—Image enhancement or correction, e.g. subtraction or averaging techniques, e.g. improvement of signal-to-noise ratio and resolution
- G01R33/561—Image enhancement or correction, e.g. subtraction or averaging techniques, e.g. improvement of signal-to-noise ratio and resolution by reduction of the scanning time, i.e. fast acquiring systems, e.g. using echo-planar pulse sequences
- G01R33/5611—Parallel magnetic resonance imaging, e.g. sensitivity encoding [SENSE], simultaneous acquisition of spatial harmonics [SMASH], unaliasing by Fourier encoding of the overlaps using the temporal dimension [UNFOLD], k-t-broad-use linear acquisition speed-up technique [k-t-BLAST], k-t-SENSE
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10088—Magnetic resonance imaging [MRI]
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/03—Recognition of patterns in medical or anatomical images
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/10—Recognition assisted with metadata
Definitions
- the content of the present disclosure relates to data processing technology, and specifically relates to a method of processing medical data to generate training data for a deep learning model.
- Magnetic resonance imaging (MRI) devices are equipment that requires considerable imaging time. Therefore, in the medical industry, accelerated imaging technology to shorten the imaging time of magnetic resonance imaging is very important and continues to develop.
- the premise of accelerated imaging of magnetic resonance imaging is that data is acquired in the signal domain, so-called k-space, rather than the image domain that humans see.
- an accelerated magnetic resonance image In order for an accelerated magnetic resonance image to be used in the medical field, it must contain all information about the subject being imaged, while minimizing noise that affects the interpretation of the information.
- technology has recently been developed based on artificial intelligence to restore low-quality, accelerated MRI images to a high-quality, non-accelerated state.
- effective learning of the artificial intelligence model In order to restore high quality accelerated magnetic resonance images based on artificial intelligence, effective learning of the artificial intelligence model must be a prerequisite.
- there are realistically difficult aspects of securing data for learning artificial intelligence models For example, in order to effectively learn an artificial intelligence model, not only high-quality input data but also high-quality label data corresponding to it must be secured.
- This disclosure was developed in response to the above-described background technology, and takes into account the characteristics of raw medical data to secure high-quality data required for learning a deep learning model to restore low-quality medical data to high quality.
- the purpose is to provide a method for doing so.
- a method of processing medical data for learning a deep learning model performed by a computing device, includes: separating the K-space data into a plurality of different data, taking into account the characteristics of metadata of the K-space data; And it may include combining a plurality of different data separated from the K-space data to generate input data and label data for learning a deep learning model.
- the metadata may include at least one of a number of excitations (NEX), an acceleration factor, a parallel imaging technique applied to the K-space data, or a resolution.
- the step of generating input data and label data for learning a deep learning model by combining a plurality of different data separated from the K-space data includes combining the plurality of data to generate the input data generating first preliminary data for generating and second preliminary data for generating the label data; and converting the first preliminary data and the second preliminary data into an image domain to generate the input data and the label data.
- combining the plurality of data to generate first preliminary data for generating the input data and second preliminary data for generating the label data may include adjusting a noise ratio between the plurality of data. It may include generating the first preliminary data and the second preliminary data through linear combination.
- the noise of the second preliminary data may include dependent noise that is correlated with the noise of the first preliminary data and independent noise that is not correlated with the noise of the first preliminary data.
- each of the dependent noise and the independent noise in the noise of the second preliminary data may be determined by a coefficient used for linear combination between the plurality of data.
- converting the first preliminary data and the second preliminary data to an image domain to generate the input data and the label data may include performing a Fourier transform on the first preliminary data and the second preliminary data. It may include generating the input data and the label data based on (Fourier transform).
- converting the first preliminary data and the second preliminary data into an image domain to generate the input data and the label data may include converting the first preliminary data and the second preliminary data to a neural network model. It may include generating the input data and the label data by inputting the data.
- the step of combining a plurality of different data separated from the K-space data to generate input data and label data for learning a deep learning model includes adjusting the resolution of the first preliminary data. may further include.
- adjusting the resolution of the first preliminary data may include a readout resolution of the first preliminary data such that the resolution of the first preliminary data has a value less than or equal to the resolution of the second preliminary data.
- it may include adjusting at least one of the phase resolutions.
- a computer program stored in a computer-readable storage medium When the computer program is executed on one or more processors, it performs operations to process medical data for learning a deep learning model.
- the operations include: separating the K-space data into a plurality of different data in consideration of the characteristics of metadata of the K-space data; And it may include an operation of combining a plurality of different data separated from the K-space data to generate input data and label data for learning a deep learning model.
- the device includes a processor including at least one core; a memory containing program codes executable on the processor; And it may include a network unit for acquiring k-space data and meta data of the k-space data.
- the processor separates the K-space data into a plurality of different data, considering the characteristics of the metadata of the K-space data, and combines the plurality of different data separated from the K-space data. , input data and label data for learning deep learning models can be created.
- the present disclosure can provide a method of securing high-quality data required for learning a deep learning model to restore low-quality medical data to high quality.
- FIG. 1 is a block diagram of a computing device according to an embodiment of the present disclosure.
- Figure 2 is a flowchart showing a method of processing medical data according to an embodiment of the present disclosure.
- Figure 3 is a conceptual diagram showing a medical data processing process according to an embodiment of the present disclosure.
- FIG. 4 is a conceptual diagram illustrating a process of dividing data considering the characteristics of metadata according to an embodiment of the present disclosure.
- FIG. 5 is a flowchart illustrating a process for dividing data considering the characteristics of metadata according to an embodiment of the present disclosure.
- Figure 6 is a flowchart showing a method of processing medical data according to an embodiment of the present disclosure.
- Figure 7 is a conceptual diagram showing a medical data processing process according to an embodiment of the present disclosure.
- Figure 8 is a flowchart showing a method of generating input data and label data by combining a plurality of data according to an embodiment of the present disclosure.
- the term “or” is intended to mean an inclusive “or” and not an exclusive “or.” That is, unless otherwise specified in the present disclosure or the meaning is not clear from the context, “X uses A or B” should be understood to mean one of natural implicit substitutions. For example, unless otherwise specified in the present disclosure or the meaning is not clear from the context, “X uses A or B” means that It can be interpreted as one of the cases where all B is used.
- N is a natural number
- N is a natural number
- components performing different functional roles may be distinguished as first components or second components.
- components that are substantially the same within the technical spirit of the present disclosure but must be distinguished for convenience of explanation may also be distinguished as first components or second components.
- module refers to a computer-related entity, firmware, software or part thereof, hardware or part thereof.
- the “module” or “unit” can be understood as a term referring to an independent functional unit that processes computing resources, such as a combination of software and hardware.
- the “module” or “unit” may be a unit composed of a single element, or may be a unit expressed as a combination or set of multiple elements.
- a “module” or “part” in the narrow sense is a hardware element or set of components of a computing device, an application program that performs a specific function of software, a process implemented through the execution of software, or a program. It can refer to a set of instructions for execution, etc.
- module or “unit” may refer to the computing device itself constituting the system, or an application running on the computing device.
- module or “unit” may be defined in various ways within a range understandable to those skilled in the art based on the contents of the present disclosure.
- model refers to a system implemented using mathematical concepts and language to solve a specific problem, a set of software units to solve a specific problem, or a process to solve a specific problem. It can be understood as an abstract model of a process.
- a neural network “model” may refer to an overall system implemented as a neural network that has problem-solving capabilities through learning. At this time, the neural network can have problem-solving capabilities by optimizing parameters connecting nodes or neurons through learning.
- a neural network “model” may include a single neural network or a neural network set in which multiple neural networks are combined.
- Data used in this disclosure may include “image.”
- image used in this disclosure may refer to multidimensional data composed of discrete image elements.
- image can be understood as a term referring to a digital representation of an object that can be seen by the human eye.
- image may refer to multidimensional data consisting of elements corresponding to pixels in a two-dimensional image.
- Image may refer to multidimensional data consisting of elements corresponding to voxels in a three-dimensional image.
- the term “medical image archiving and communication system” refers to the storage, processing, and processing of medical images in accordance with the DICOM (digital imaging and communications in medicine) standard. It can refer to a transmitting system.
- the “medical image storage and transmission system” is linked with digital medical imaging equipment to produce medical images such as magnetic resonance imaging (MRI) and computed tomography (CT) images, and other digital medical images. It can be saved according to communication standards.
- the “medical image storage and transmission system” can transmit medical images to terminals inside and outside the hospital through a communication network. At this time, meta information such as reading results and medical records may be added to the medical image.
- K-space used in the present disclosure can be understood as an array of numbers representing the spatial frequency of a magnetic resonance image.
- K-space can be understood as a frequency space corresponding to a three-dimensional space corresponding to magnetic resonance space coordinates.
- Frier transform used in this disclosure can be understood as a computational medium that allows explaining the relationship between the time domain and the frequency domain.
- “Fourier transform” used in the present disclosure can be understood as a broad concept representing a computational process for mutual transformation between the time domain and the frequency domain. Therefore, the "Fourier transform” used in this disclosure is a concept that encompasses both the Fourier transform in the narrow sense, which decomposes a signal in the time domain into the frequency domain, and the inverse Fourier transform, which transforms the signal in the frequency domain into the time domain. It can be understood as
- FIG. 1 is a block diagram of a computing device according to an embodiment of the present disclosure.
- the computing device 100 may be a hardware device or part of a hardware device that performs comprehensive processing and calculation of data, or may be a software-based computing environment connected to a communication network.
- the computing device 100 may be a server that performs intensive data processing functions and shares resources, or it may be a client that shares resources through interaction with the server.
- the computing device 100 may be a cloud system in which a plurality of servers and clients interact to comprehensively process data. Since the above description is only an example related to the type of computing device 100, the type of computing device 100 may be configured in various ways within a range understandable to those skilled in the art based on the contents of the present disclosure.
- a computing device 100 may include a processor 110, a memory 120, and a network unit 130. there is. However, since FIG. 1 is only an example, the computing device 100 may include other components for implementing a computing environment. Additionally, only some of the configurations disclosed above may be included in computing device 100.
- the processor 110 may be understood as a structural unit including hardware and/or software for performing computing operations.
- the processor 110 may read a computer program and perform data processing for machine learning.
- the processor 110 may process computational processes such as processing input data for machine learning, extracting features for machine learning, and calculating errors based on backpropagation.
- the processor 110 for performing such data processing includes a central processing unit (CPU), a general purpose graphics processing unit (GPGPU), a tensor processing unit (TPU), and a custom processing unit (TPU). It may include a semiconductor (ASIC: application specific integrated circuit), or a field programmable gate array (FPGA: field programmable gate array). Since the type of processor 110 described above is only an example, the type of processor 110 may be configured in various ways within a range understandable to those skilled in the art based on the contents of the present disclosure.
- the processor 110 may generate training data for a deep learning model to restore low-quality medical data to high quality.
- the processor 110 may use metadata associated with medical data to generate input data and label data for learning a deep learning model.
- the processor 110 may consider the characteristics of metadata of medical data and generate input data and label data for learning a deep learning model from medical data.
- the medical data may include accelerated imaging or accelerated simulation (simulation) magnetic resonance signals or images.
- the deep learning model may include a neural network model that restores accelerated imaging or accelerated simulated magnetic resonance images to normal imaging conditions.
- the deep learning model may include a neural network model that generates output to reduce noise or improve resolution of accelerated imaging or accelerated simulated magnetic resonance images.
- the neural network model may include U-NET, which is based on a convolutional neural network. Since the above-described type of neural network is only an example, it can be configured in a variety of ways that can be understood by those skilled in the computer vision field.
- accelerated imaging can be understood as an imaging technique that shortens imaging time by reducing the number of excitations (NEX) for magnetic resonance signals compared to general imaging.
- the number of excitations can be understood as the number of repetitions when repeatedly acquiring lines of a magnetic resonance signal in the K-space domain. Therefore, as the number of excitations increases, the capturing time of the magnetic resonance image may increase proportionally. That is, when the number of excitations is reduced when taking a magnetic resonance image, accelerated imaging with a shortened magnetic resonance image taking time can be implemented.
- accelerated imaging can be understood as an imaging technique that obtains an image with relatively low resolution by obtaining a narrower range of signals in the phase encoding direction in the K-space domain.
- accelerated imaging of the present disclosure can be understood as a imaging technique that reduces phase resolution compared to general imaging.
- Phase resolution can be understood as the number of lines sampled in the phase encoding direction in the K-space domain divided by a preset reference value. Therefore, as the phase resolution increases, the capturing time of the magnetic resonance image may increase proportionally. That is, when the phase resolution is reduced when capturing a magnetic resonance image, accelerated imaging with a shortened magnetic resonance image capturing time can be implemented.
- accelerated shooting can be understood as a shooting technique that shortens shooting time by increasing the acceleration factor compared to regular shooting.
- the acceleration index is a term used in parallel imaging techniques, and can be understood as the number of signal lines fully sampled in K-space divided by the number of signal lines sampled through imaging.
- an acceleration index of 2 can be understood as acquiring half the number of signal lines compared to the number of fully sampled signal lines when acquiring a line by sampling a magnetic resonance signal in the phase encoding direction. Therefore, as the acceleration index increases, the capturing time of the magnetic resonance image may decrease proportionally. That is, if the acceleration index is increased when taking a magnetic resonance image, accelerated imaging with a shortened magnetic resonance image taking time can be implemented.
- accelerated imaging can be understood as an imaging technique that generates a magnetic resonance image by acquiring a sub-sampled magnetic resonance signal.
- subsampling can be understood as an operation of sampling a magnetic resonance signal at a sampling rate lower than the Nyquist sampling rate.
- the medical data of the present disclosure may be images obtained by sampling magnetic resonance signals at a sampling rate lower than the Nyquist sampling rate.
- Accelerated simulation of the present disclosure can be understood as a computational technique for under sampling K-space data generated through regular or accelerated imaging.
- undersampling can be understood as a method of processing magnetic resonance signals at a lower sampling rate based on the K-space data to be processed.
- accelerated simulation may include computational techniques to generate subsampled K-space data based on fully sampled K-space data.
- accelerated simulations may include computational techniques to sample magnetic resonance signals using subsampled K-space data at a lower sampling rate.
- the accelerated simulation technique described above may be applied as is. Acceleration simulation may be performed by the processor 110 of the present disclosure, or may be performed through a separate external system.
- the processor 110 may separate the K-space data into a plurality of different data based on the metadata of the K-space data.
- the processor 110 may analyze the metadata of the K-space data and divide the K-space data into a plurality of different data based on the encoding line of the K-space data.
- the processor 110 may consider the characteristics of metadata of K-space data and generate a plurality of different data differentiated based on encoding lines from K-space data.
- the K-space data may correspond to a magnetic resonance signal or image of the K-space domain that has been accelerated or simulated.
- the processor 110 may determine what characteristics the metadata of K-space data has. Additionally, the processor 110 may select an appropriate division technique based on the above-described determination result to generate a plurality of different data from the K-space data. In other words, the processor 110 may identify a partitioning technique for separating the K-space data into a plurality of different data according to the characteristics of the metadata of the K-space data. Additionally, the processor 110 may generate a plurality of different data from K-space data using the previously identified segmentation technique. For example, the processor 110 may select a partitioning technique that matches the characteristics of the metadata of the K-space data that is currently being processed from among partitioning techniques previously classified according to the characteristics of the metadata.
- the metadata may include at least one of the number of excitations, acceleration index, parallel imaging technique applied to the K-space data, or resolution.
- the processor 110 may use a segmentation technique that matches the characteristics of the metadata of the K-space data currently being processed to generate a plurality of different data distinguished based on the encoding line from the K-space data.
- the processor 110 can consider what acceleration characteristics the K-space data has and route an appropriate partitioning technique that matches the characteristics of the K-space data. Through this routing and data processing, the processor 110 can prepare basic data for generating input data and label data optimized for learning a deep learning model from K-space data.
- the processor 110 may generate training data for a deep learning model based on a plurality of different data generated from K-space data.
- the processor 110 may combine a plurality of different data separated from the K-space data to generate input data and label data for learning a deep learning model.
- the processor 110 may combine a plurality of different data to generate preliminary data to prepare input data for learning a deep learning model.
- the processor 110 may combine a plurality of different pieces of data to generate preliminary data for preparing label data.
- the processor 110 performs an operation to generate preliminary data for the input data and preliminary data for the label data in consideration of the combination ratio between the plurality of data. Can be performed individually.
- the processor 110 may generate input data and label data for learning a deep learning model based on each preliminary data.
- the processor 110 can generate input data for learning a deep learning model and high-quality label data corresponding to the input data from single K-space data, considering the inherent characteristics of the K-space data. Through this data processing process, the processor 110 can effectively build learning data for a deep learning model for restoring accelerated imaging or accelerated simulated magnetic resonance images.
- the memory 120 may be understood as a structural unit including hardware and/or software for storing and managing data processed in the computing device 100. That is, the memory 120 can store any type of data generated or determined by the processor 110 and any type of data received by the network unit 130.
- the memory 120 may be a flash memory type, hard disk type, multimedia card micro type, card type memory, or random access memory (RAM). ), SRAM (static random access memory), ROM (read-only memory), EEPROM (electrically erasable programmable read-only memory), PROM (programmable read-only memory), magnetic memory , may include at least one type of storage medium among a magnetic disk and an optical disk.
- the memory 120 may include a database system that controls and manages data in a predetermined system. Since the type of memory 120 described above is only an example, the type of memory 120 may be configured in various ways within a range understandable to those skilled in the art based on the contents of the present disclosure.
- the memory 120 can manage data necessary for the processor 110 to perform operations, a combination of data, and program code executable on the processor 110 by structuring and organizing them.
- the memory 120 may store medical data received through the network unit 130, which will be described later.
- the memory 120 is a program code that operates to process medical data to generate training data for a neural network model, and a program code that operates the processor 110 to generate image data based on feature interpretation (or inference) of the neural network model. And K-space data, image data, etc. generated as the program code is executed can be stored.
- the network unit 130 may be understood as a structural unit that transmits and receives data through any type of known wired or wireless communication system.
- the network unit 130 is a local area network (LAN), wideband code division multiple access (WCDMA), long term evolution (LTE), and WiBro (wireless).
- broadband internet 5th generation mobile communication (5G), ultra wide-band wireless communication, ZigBee, radio frequency (RF) communication, wireless LAN, wireless fidelity ), data transmission and reception can be performed using a wired or wireless communication system such as near field communication (NFC), or Bluetooth. Since the above-described communication systems are only examples, the wired and wireless communication systems for data transmission and reception of the network unit 130 may be applied in various ways other than the above-described examples.
- the network unit 130 may receive data necessary for the processor 110 to perform calculations through wired or wireless communication with any system or client. Additionally, the network unit 130 may transmit data generated through the calculation of the processor 110 through wired or wireless communication with any system or any client. For example, the network unit 130 may receive medical data through communication with a medical image storage and transmission system, a cloud server that performs tasks such as standardization of medical data, or a computing device. The network unit 130 may transmit learning data generated through calculations of the processor 110, image data corresponding to the output of the neural network model, etc. through communication with the above-described system, server, or computing device.
- Figure 2 is a flowchart showing a method of processing medical data according to an embodiment of the present disclosure.
- the computing device 100 may acquire K-space data and metadata of the K-space data (S100).
- “acquisition” can be understood to refer not only to receiving data through a wireless communication network with an external terminal, device, or system, but also to generating or receiving data in an on-device form.
- the computing device 100 may receive K-space data and metadata of the K-space data through cloud communication with a medical image storage and transmission system or a magnetic resonance imaging device.
- the computing device 100 may be mounted on a magnetic resonance imaging device and directly obtain K-space data and metadata of the K-space data.
- the metadata of the K-space data may be metadata related to accelerated shooting of the K-space data.
- Metadata of K-space data may include at least one of the number of excitations, acceleration exponent, parallel imaging technique applied to K-space data, or resolution.
- the computing device 100 may separate the K-space data into a plurality of different data based on the encoding line of the K-space data, considering the characteristics of the metadata acquired through step S100 (S200).
- the computing device 100 may analyze numerical values representing characteristics of metadata and generate a plurality of data that are distinguished from each other based on encoding lines from K-space data. For example, the computing device 100 may use a segmentation technique appropriate to apply to the K-space data acquired through step S100, based on at least one numerical value of the number of excitations, acceleration exponent, parallel imaging technique, or resolution. can be identified.
- identification can be understood as a task of matching an appropriate segmentation technique among pre-classified segmentation techniques according to a combination of at least one numerical value among the number of excitations, acceleration exponent, parallel imaging technique, or resolution.
- the computing device 100 may use the identified segmentation technique to generate a plurality of data that are distinguished from each other based on encoding lines from K-space data.
- the plurality of data can be understood as basic data used to generate input data and label data for learning a deep learning model.
- Figure 3 is a conceptual diagram showing a medical data processing process according to an embodiment of the present disclosure.
- Figure 4 is a conceptual diagram showing a process of dividing data considering the characteristics of metadata according to an embodiment of the present disclosure.
- the processor 110 of the computing device 100 based on K-space data 210 and metadata 220 of the K-space data 210, A plurality of data 300 can be generated to be used to prepare training data for a deep learning model for restoration of medical data.
- the processor 110 may identify a partitioning technique for separating the K-space data 210 into a plurality of different data 300 based on the characteristics of the metadata 220.
- the characteristics of the metadata 220 may be parameters related to accelerated imaging of the K-space data 210, or attribute values indicating an accelerated imaging technique or restoration technique.
- the processor 110 may use the identified segmentation technique to generate a plurality of data 300 in which noise is independent from the K-space data 210 based on the encoding line.
- the encoding line can be understood as a set of frequency encoding or phase encoding that forms the dimension of K-space data.
- the processor 110 may determine a combination of numerical values representing characteristics of the metadata 220.
- the processor 110 may determine a partitioning technique matching the combination as a partitioning technique used to generate a plurality of data 300 in which noise is independent of each other.
- the processor 110 may select a segmentation technique that matches the combination of numerical values among pre-classified segmentation techniques based on a combination of numerical values representing characteristics of the metadata 200. That is, the processor 110 can individually consider how the K-space data was captured and appropriately use a segmentation technique appropriate for the K-space data. In other words, the processor 110 can determine a segmentation technique suitable for the K-space data by considering the characteristics of metadata for the K-space data that can be used as a clinical standard. This data processing process ensures that no matter what K-space data is secured, high-quality data with noise independent of each other can be secured through techniques optimized for the K-space data itself.
- the segmentation technique may include a first segmentation technique 10 that generates a plurality of data with mutually independent noise from K-space data, based on the order of accelerated shooting according to the number of excitations. You can. Additionally, the division technique may include a second division technique 20 that generates a plurality of data with independent noise from K-space data based on the interval of the phase encoding line. For example, the first division technique 10 determines whether it is the first or second captured data according to the number of excitations, and divides the K-space data according to the shooting order to generate a plurality of data. Techniques may be included.
- the second division technique 20 distinguishes the phase encoding lines of the K-space data by setting the spacing of the phase encoding lines to 2, and generates a plurality of data whose noise is mutually independent from the K-space data based on the differentiated phase encoding lines. It may include a segmentation technique that generates . That is, the second division technique 20 may include a technique of determining whether the phase encoding line is odd or even and dividing the K-space data according to this pattern to generate a plurality of data. In addition to the above-described example where the spacing of the phase encoding lines is 2, the second division technique 20 may also include a case where the phase encoding spacing is 2 or more.
- the second division technique 20 distinguishes the phase encoding lines of K-space data by setting the spacing of the phase encoding lines to n (n is a natural number greater than 2), and divides the phase encoding lines into K-space based on the differentiated phase encoding lines. It may include a segmentation technique that generates a plurality of data with mutually independent noise from data.
- the first division technique 10 and the second division technique 20 described above may be classified in advance by matching individual combinations of numerical values representing characteristics of metadata. Accordingly, the processor 110 uses at least one of the pre-classified first segmentation technique 10 or the second segmentation technique 20 based on a combination of numerical values representing the characteristics of the meta data 220 to reduce noise. It can be determined by the partitioning technique used to generate a plurality of mutually independent data 300. Specifically, referring to FIG. 4, when the combination of numerical values representing the characteristics of the metadata 220 is [number of excitations A, acceleration index B], the processor 110 uses the first segmentation technique 10, or the second segmentation technique 10.
- the first partition technique 10 matching [number of excitations A, acceleration index B] is used to partition the K-space data 210. It can be decided by a partitioning technique to do this.
- the 2-1 division technique 21 included in the second division technique 20 may correspond to a division technique in which the interval between phase encoding lines is 2.
- the 2-2 division technique 22 included in the second division technique 20 may correspond to a division technique in which the spacing of phase encoding lines is 3.
- segmentation technique may be additionally defined in a range understandable to those skilled in the art based on the contents of the present disclosure in addition to the above-described example.
- the processor 110 may generate a plurality of data 300 in which noise is mutually independent from the K-space data 210 based on the encoding line, using at least one of the pre-classified segmentation techniques. For example, referring to FIG. 4, according to a combination of numerical values representing characteristics of the meta data 220, the processor 110 uses the first partitioning technique 10 to divide the K-space data 210 into 2 First divided data 310 and second divided data 320 can be generated by dividing the data into pieces. At this time, the signal strength of the first divided data 310 and the second divided data 320 corresponds to each other, but at least one of the noise intensity or distribution may be different and have an independent relationship.
- the processor 110 generates a plurality of data 300 in which at least one of the intensity or distribution of noise is independent from the K-space data 210 through a partitioning technique matching the characteristics of the metadata 220. By doing so, it is possible to efficiently construct a high-quality data set that serves as the basis for generating input data and label data, which will be described later.
- FIG. 5 is a flowchart illustrating a process for dividing data considering the characteristics of metadata according to an embodiment of the present disclosure. Detailed content matching the description of FIGS. 3 and 4 for each step in FIG. 5 will be omitted.
- the processor 110 of the computing device 100 may determine a combination of numerical values representing characteristics of metadata (S210). Additionally, the processor 110 may identify a segmentation technique that matches the combination among the previously classified segmentation techniques (S220). For example, as shown in FIG. 5 , the first segmentation technique 10 and the second segmentation technique 20 may be classified in advance according to combinations of numerical values representing characteristics of metadata. At this time, if the combination of characteristic values of the metadata to be processed is [number of excitations A1, acceleration index B1], the processor 110 uses the first partitioning technique 10 matching [number of excitations A1, acceleration index B1] or At least one of the second segmentation techniques 20 may be determined as the segmentation technique to be applied to the processing target.
- the processor 110 may determine the partitioning technique identified in step S220 as the partitioning technique to be used to separate the plurality of data from the K-space data (S230). Then, the processor 110 may use the determined segmentation technique to generate a plurality of data in which noise is independent of each other (S240).
- Figure 6 is a flowchart showing a method of processing medical data according to an embodiment of the present disclosure.
- the computing device 100 may separate the K-space data into a plurality of different data in consideration of the characteristics of the metadata of the K-space data (S310 ). Since the specific content related to step S310 matches the step S120 of FIG. 2 described above, detailed description will be omitted.
- the computing device 100 may generate input data and label data for learning a deep learning model by combining a plurality of different data separated from the K-space data (S320).
- the plurality of different data can be understood as a plurality of data in which at least one of the intensity or distribution of noise is independent.
- the computing device 100 may combine a plurality of different data to prepare preliminary data to be used to generate input data and label data for learning a deep learning model.
- the computing device 100 may convert preliminary data into an image domain and generate input data and label data to be used for learning a deep learning model. That is, the computing device 100 prepares preliminary data for generating input data and label data for learning a deep learning model using the basic data prepared through step S310, and input data and label data through domain conversion of the preliminary data. Label data can be created.
- Figure 7 is a conceptual diagram showing a medical data processing process according to an embodiment of the present disclosure.
- the processor 110 of the computing device 100 combines a plurality of data 300 with mutually independent noise to generate input data 510.
- Second preliminary data 420 for generating first preliminary data 410 and label data 520 may be generated.
- the processor 110 may include first preliminary data 410 for generating input data 510 and second preliminary data 410 for generating label data 520 through linear combination that adjusts the noise ratio between a plurality of data.
- Preliminary data 420 may be generated respectively. Specifically, it is assumed that through the processing process as shown in the example of FIG. 4, two pieces of data with the same signal intensity but independent noise intensity and distribution are generated.
- the processor 110 may generate first preliminary data 410 by linearly combining the two data by applying a linear combination coefficient of 1 to one of the two data and a linear combination coefficient of 0 to the other one. That is, the first preliminary data 410 may correspond to data to which a linear combination coefficient of 1 has been applied.
- the processor 110 linearly combines the two data by applying a linear combination coefficient of 1/3 to the data corresponding to the first preliminary data 410 and a linear combination coefficient of 2/3 to the remaining one to form the second preliminary data.
- Data 420 may be generated. That is, the second preliminary data 420 may be generated as data including noise with noise reduced to 1/3 compared to the first preliminary data 410.
- the processor 110 individually performs the purpose of generating input data 510 or label data 520 based on a plurality of data 300 in which noise is independent of each other. It is possible to generate preliminary data 410 and 420 in which the noise ratio is adjusted in various ways according to the linear combination coefficient.
- the noise of the second preliminary data 420 has a dependent noise that has a correlation with the noise of the first preliminary data 410 and has no correlation with the noise of the first preliminary data 410. It may contain independent noise. Specifically, the dependent noise may correspond to noise obtained by applying a linear combination coefficient of 1/3 to the noise of data corresponding to the first preliminary data 410 among the plurality of data. Additionally, the independent noise may correspond to noise obtained by applying a linear combination coefficient of 2/3 to the noise of the remaining data that does not correspond to the first preliminary data 410.
- the ratio of linear combination coefficients in the preceding examples is only an example, and the ratio of linear combination coefficients can be adjusted in various ways to suit the purpose of generating input data and label data.
- the processor 110 may convert the first preliminary data 410 and the second preliminary data 420 into an image domain to generate input data 510 and label data 520, respectively.
- the processor 110 may generate the input data 510 and the label data 520 based on Fourier transform of the first preliminary data 410 and the second preliminary data 420.
- the processor 110 inputs the first preliminary data 410 and the second preliminary data 420 into a neural network model that converts the K-space domain into an image domain, creating input data 510 and label data 520, respectively.
- You can also create The data processing process described in Figure 7 uses data with mutually independent noise to generate high-quality data needed for each purpose, such as input or label for learning a deep learning model, as quickly and accurately as desired.
- Figure 8 is a flowchart showing a method of generating input data and label data by combining a plurality of data according to an embodiment of the present disclosure. Detailed content that matches the description of FIG. 7 for each step of FIG. 8 will be omitted.
- the computing device 100 may apply a parallel imaging technique to a plurality of different data with mutually independent noise (S410).
- Parallel imaging technique can be understood as an image processing technique to restore missing information in accelerated magnetic resonance images.
- the computing device 100 may restore a plurality of data by reconstructing autocalibrating signal (ACS) lines of a plurality of different data.
- the computing device 100 may restore a plurality of data by reconstructing magnetic resonance signal lines that constitute a plurality of different data.
- the computing device 100 may restore a plurality of data using a sensitivity map based on sensitivity information for each channel of the coil to obtain K-space data.
- various parallel imaging techniques can be applied within a range understandable to those skilled in the art based on the contents of the present disclosure. If the plurality of data is data restored through a parallel imaging technique, step S410 may not be performed.
- the processor 110 may generate preliminary data for generating training data for a deep learning model that restores the quality of medical data by linearly combining a plurality of data with mutually independent noise (S420). At this time, the plurality of data can be understood as data restored through a parallel imaging technique.
- the processor 110 may adjust the resolution of preliminary data to create input data to be used for learning a deep learning model (S430). For example, the processor 110 may perform one of the basic resolution or phase resolution of the first preliminary data so that the resolution of the first preliminary data related to the input data has a value less than or equal to the resolution of the second preliminary data related to the label data. At least one can be adjusted.
- the basic resolution is the number of pixels in the readout direction and can be understood as the relative size of data existing in the frequency encoding direction.
- phase resolution is the number of pixels in the phase encoding direction and can be understood as the relative size of data existing in the phase encoding direction.
- the range in which the resolution is adjusted may be determined according to a value preset by the user. Therefore, the resolution of the input data to be used for learning a deep learning model can be adjusted in various ways within a lower range than the resolution of the label data.
- the processor 110 may generate input data and label data by converting the domain of the preliminary data whose resolution has been adjusted through step S430 (S440). Additionally, the processor 110 can use the input data and label data as learning data for a deep learning model to restore the quality of medical data.
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Public Health (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Epidemiology (AREA)
- Primary Health Care (AREA)
- Radiology & Medical Imaging (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Pathology (AREA)
- High Energy & Nuclear Physics (AREA)
- Heart & Thoracic Surgery (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- Veterinary Medicine (AREA)
- Condensed Matter Physics & Semiconductors (AREA)
- Signal Processing (AREA)
- Image Processing (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
Abstract
Description
Claims (12)
- 적어도 하나의 프로세서를 포함하는 컴퓨팅 장치에 의해 수행되는, 딥러닝 모델의 학습을 위한 의료 데이터의 처리 방법으로서,케이-스페이스 데이터의 메타 데이터의 특성을 고려하여, 상기 케이-스페이스 데이터를 서로 다른 복수의 데이터로 분리하는 단계; 및상기 케이-스페이스 데이터로부터 분리된 서로 다른 복수의 데이터를 조합하여, 딥러닝 모델의 학습을 위한 입력 데이터 및 라벨 데이터를 생성하는 단계;를 포함하는,방법.
- 제 1 항에 있어서,상기 메타 데이터는,여기 횟수(NEX), 가속화 지수(acceleration factor), 상기 케이-스페이스 데이터에 적용된 병렬 영상 기법, 또는 레졸루션(resolution) 중 적어도 하나를 포함하는,방법.
- 제 1 항에 있어서,상기 케이-스페이스 데이터로부터 분리된 서로 다른 복수의 데이터를 조합하여, 딥러닝 모델의 학습을 위한 입력 데이터 및 라벨 데이터를 생성하는 단계는,상기 복수의 데이터를 결합하여, 상기 입력 데이터를 생성하기 위한 제 1 예비 데이터 및 상기 라벨 데이터를 생성하기 위한 제 2 예비 데이터를 생성하는 단계; 및상기 제 1 예비 데이터 및 상기 제 2 예비 데이터를 이미지 도메인으로 변환하여, 상기 입력 데이터 및 상기 라벨 데이터를 생성하는 단계;를 포함하는,방법.
- 제 3 항에 있어서,상기 복수의 데이터를 결합하여, 상기 입력 데이터를 생성하기 위한 제 1 예비 데이터 및 상기 라벨 데이터를 생성하기 위한 제 2 예비 데이터를 생성하는 단계는,상기 복수의 데이터 간의 노이즈 비율을 조정하는 선형 결합을 통해, 상기 제 1 예비 데이터 및 상기 제 2 예비 데이터를 생성하는 단계;를 포함하는,방법.
- 제 4 항에 있어서,상기 제 2 예비 데이터의 노이즈는,상기 제 1 예비 데이터의 노이즈와 상관 관계를 갖는 종속 노이즈 및 상기 제 1 예비 데이터의 노이즈와 상관 관계를 갖지 않는 독립 노이즈를 포함하는,방법.
- 제 5 항에 있어서,상기 제 2 예비 데이터의 노이즈에서 상기 종속 노이즈와 상기 독립 노이즈 각각은,상기 복수의 데이터 간의 선형 결합을 위해 사용되는 계수에 의해 결정되는,방법.
- 제 3 항에 있어서,상기 제 1 예비 데이터 및 상기 제 2 예비 데이터를 이미지 도메인으로 변환하여, 상기 입력 데이터 및 상기 라벨 데이터를 생성하는 단계는,상기 제 1 예비 데이터 및 상기 제 2 예비 데이터에 대한 푸리에 변환(fourier transform)을 기초로, 상기 입력 데이터 및 상기 라벨 데이터를 생성하는 단계;를 포함하는,방법.
- 제 3 항에 있어서,상기 제 1 예비 데이터 및 상기 제 2 예비 데이터를 이미지 도메인으로 변환하여, 상기 입력 데이터 및 상기 라벨 데이터를 생성하는 단계는,상기 제 1 예비 데이터 및 상기 제 2 예비 데이터를 신경망 모델에 입력하여, 상기 입력 데이터 및 상기 라벨 데이터를 생성하는 단계;를 포함하는,방법.
- 제 3 항에 있어서,상기 케이-스페이스 데이터로부터 분리된 서로 다른 복수의 데이터를 조합하여, 딥러닝 모델의 학습을 위한 입력 데이터 및 라벨 데이터를 생성하는 단계는,상기 제 1 예비 데이터의 레졸루션을 조정하는 단계;를 더 포함하는,방법.
- 제 9 항에 있어서,상기 제 1 예비 데이터의 레졸루션을 조정하는 단계는,상기 제 1 예비 데이터의 레졸루션이 상기 제 2 예비 데이터의 레졸루션 이하의 값을 갖도록, 상기 제 1 예비 데이터의 베이직(basic) 레졸루션 혹은 위상 레졸루션 중 적어도 하나를 조정하는 단계;를 포함하는,방법.
- 컴퓨터 판독가능 저장 매체 저장된 컴퓨터 프로그램(program)으로서, 상기 컴퓨터 프로그램은 하나 이상의 프로세서(processor)에서 실행되는 경우, 딥러닝 모델의 학습을 위해 의료 데이터를 처리하는 동작들을 수행하도록 하며,상기 동작들은,케이-스페이스 데이터의 메타 데이터의 특성을 고려하여, 상기 케이-스페이스 데이터를 서로 다른 복수의 데이터로 분리하는 동작; 및상기 케이-스페이스 데이터로부터 분리된 서로 다른 복수의 데이터를 조합하여, 딥러닝 모델의 학습을 위한 입력 데이터 및 라벨 데이터를 생성하는 동작;을 포함하는,컴퓨터 프로그램.
- 딥러닝 모델의 학습을 위해 의료 데이터를 처리하는 컴퓨팅 장치로서,적어도 하나의 코어(core)를 포함하는 프로세서(processor);상기 프로세서에서 실행 가능한 프로그램 코드(code)들을 포함하는 메모리(memory); 및케이-스페이스(k-space) 데이터 및 상기 케이-스페이스 데이터의 메타(meta) 데이터를 획득하기 위한 네트워크부(network unit);를 포함하고,상기 프로세서는,케이-스페이스 데이터의 메타 데이터의 특성을 고려하여, 상기 케이-스페이스 데이터를 서로 다른 복수의 데이터로 분리하고,상기 케이-스페이스 데이터로부터 분리된 서로 다른 복수의 데이터를 조합하여, 딥러닝 모델의 학습을 위한 입력 데이터 및 라벨 데이터를 생성하는,장치.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US18/721,804 US20250086942A1 (en) | 2022-06-22 | 2023-06-20 | Method, program, and device for processing medical data for training of deep learning model |
| EP23827486.4A EP4506953A4 (en) | 2022-06-22 | 2023-06-20 | METHOD, PROGRAM, AND DEVICE FOR PROCESSING MEDICAL DATA FOR TRAINING A DEEP LEARNING MODEL |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| KR10-2022-0076366 | 2022-06-22 | ||
| KR1020220076366A KR102546620B1 (ko) | 2022-06-22 | 2022-06-22 | 딥러닝 모델의 학습을 위한 의료 데이터의 처리 방법, 프로그램 및 장치 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2023249372A1 true WO2023249372A1 (ko) | 2023-12-28 |
Family
ID=86988945
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/KR2023/008528 Ceased WO2023249372A1 (ko) | 2022-06-22 | 2023-06-20 | 딥러닝 모델의 학습을 위한 의료 데이터의 처리 방법, 프로그램 및 장치 |
Country Status (4)
| Country | Link |
|---|---|
| US (1) | US20250086942A1 (ko) |
| EP (1) | EP4506953A4 (ko) |
| KR (2) | KR102546620B1 (ko) |
| WO (1) | WO2023249372A1 (ko) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR102859458B1 (ko) * | 2024-10-28 | 2025-09-19 | 주식회사 에어스메디컬 | 딥 러닝에 모델에 기반한 자기 공명 영상을 획득하는 장치 및 그 제어 방법 |
| US12584985B1 (en) | 2024-10-28 | 2026-03-24 | Airs Medical Inc. | Apparatus for obtaining magnetic resonance images base on deep learning model and method of controlling the same |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR102199494B1 (ko) * | 2019-07-15 | 2021-01-06 | 주식회사 딥노이드 | 의료 영상 관리 장치 및 방법 |
| KR20220082302A (ko) * | 2020-12-10 | 2022-06-17 | 주식회사 에어스 메디컬 | 리스케일링과 인공신경망을 적용한 자기 공명 영상 처리 장치 및 그 방법 |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10573031B2 (en) * | 2017-12-06 | 2020-02-25 | Siemens Healthcare Gmbh | Magnetic resonance image reconstruction with deep reinforcement learning |
| US11125844B2 (en) * | 2018-07-06 | 2021-09-21 | The Board Of Trustees Of The Leland Stanford Junior University | Deep learning based methods to accelerate multi-spectral imaging |
| CN109658471B (zh) * | 2018-12-20 | 2023-07-25 | 上海联影医疗科技股份有限公司 | 一种医学图像重建方法和系统 |
| KR102328214B1 (ko) | 2021-05-17 | 2021-11-19 | (주)제이엘케이 | 의료 데이터의 전처리를 통한 의료 데이터베이스 구축 시스템 및 동작 방법 |
-
2022
- 2022-06-22 KR KR1020220076366A patent/KR102546620B1/ko active Active
-
2023
- 2023-06-16 KR KR1020230077273A patent/KR20230175117A/ko active Pending
- 2023-06-20 EP EP23827486.4A patent/EP4506953A4/en active Pending
- 2023-06-20 WO PCT/KR2023/008528 patent/WO2023249372A1/ko not_active Ceased
- 2023-06-20 US US18/721,804 patent/US20250086942A1/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR102199494B1 (ko) * | 2019-07-15 | 2021-01-06 | 주식회사 딥노이드 | 의료 영상 관리 장치 및 방법 |
| KR20220082302A (ko) * | 2020-12-10 | 2022-06-17 | 주식회사 에어스 메디컬 | 리스케일링과 인공신경망을 적용한 자기 공명 영상 처리 장치 및 그 방법 |
Non-Patent Citations (4)
| Title |
|---|
| OH CHANGHEUN, KIM DONGCHAN, CHUNG JUN‐YOUNG, HAN YEJI, PARK HYUNWOOK: "A k‐space‐to‐image reconstruction network for MRI using recurrent neural network", MEDICAL PHYSICS., AIP, MELVILLE, NY., US, vol. 48, no. 1, 1 January 2021 (2021-01-01), US , pages 193 - 203, XP093118690, ISSN: 0094-2405, DOI: 10.1002/mp.14566 * |
| RAMANARAYANAN SRIPRABHA, MURUGESAN BALAMURALI, KEERTHI RAM, SIVAPRAKASAM MOHANASANKAR: "MAC-ReconNet: A Multiple Acquisition Context based Convolutional Neural Network for MR Image Reconstruction using Dynamic Weight Prediction", ARXIV (CORNELL UNIVERSITY), CORNELL UNIVERSITY LIBRARY, ARXIV.ORG, ITHACA, 9 November 2021 (2021-11-09), Ithaca, XP093118680, [retrieved on 20240112], DOI: 10.48550/arxiv.2111.05055 * |
| See also references of EP4506953A4 * |
| ZALAN FABIAN, HECKEL REINHARD, SOLTANOLKOTABI MAHDI: "Data augmentation for deep learning based accelerated MRI reconstruction with limited data", ICONFERENCE, ASSOCIATION FOR COMPUTING MACHINERY NEW YORK, NY, UNITED STATES, vol. 139, 28 June 2021 (2021-06-28), pages 3057 - 3067, XP093118686, ISSN: 2331-8422 * |
Also Published As
| Publication number | Publication date |
|---|---|
| US20250086942A1 (en) | 2025-03-13 |
| KR20230175117A (ko) | 2023-12-29 |
| EP4506953A4 (en) | 2025-08-27 |
| EP4506953A1 (en) | 2025-02-12 |
| KR102546620B1 (ko) | 2023-06-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111178542B (zh) | 基于机器学习建模的系统和方法 | |
| CN110766769B (zh) | 一种磁共振图像重建方法、装置、设备和介质 | |
| CN109410253B (zh) | 用于生成信息的方法、装置、电子设备和计算机可读介质 | |
| WO2023249372A1 (ko) | 딥러닝 모델의 학습을 위한 의료 데이터의 처리 방법, 프로그램 및 장치 | |
| WO2024039043A1 (ko) | 학습 데이터 생성 방법, 컴퓨터 프로그램 및 장치 | |
| US20230135995A1 (en) | Multi-slice mri data processing using deep learning techniques | |
| WO2023229384A1 (ko) | 학습 데이터 생성 방법, 컴퓨터 프로그램 및 장치 | |
| CN110738235A (zh) | 肺结核判定方法、装置、计算机设备及存储介质 | |
| WO2020060196A1 (ko) | 3차원 영상 재구성 장치 및 그 방법 | |
| CN113610911A (zh) | 深度预测模型的训练方法及装置、介质和电子设备 | |
| WO2021137454A1 (ko) | 인공지능 기반의 사용자 의료정보 분석 방법 및 시스템 | |
| KR102683611B1 (ko) | Mri 이미지의 노이즈 제거를 위한 후처리 장치의 학습방법 및 이를 위한 컴퓨팅 장치 | |
| WO2023249402A1 (ko) | 딥러닝 모델의 학습을 위한 의료 데이터의 처리 방법, 프로그램 및 장치 | |
| Yu et al. | Universal generative modeling in dual domains for dynamic MRI | |
| CN116309901B (zh) | 伪影去除方法、装置、电子设备及存储介质 | |
| WO2023048437A1 (ko) | 의료 데이터를 기반으로 하는 딥러닝 모델의 학습 및 추론 방법, 프로그램 및 장치 | |
| WO2023177151A1 (ko) | 3차원 의료 데이터의 처리 방법, 프로그램 및 장치 | |
| WO2024053828A1 (ko) | 인공 신경망 기반 해상도 향상 방법, 프로그램 및 장치 | |
| WO2022092670A1 (ko) | 뇌 피질 영역의 두께를 분석하는 방법 | |
| WO2020230972A1 (ko) | 훈련된 심층 신경망 모델의 재현 성능을 개선하는 방법 및 이를 이용한 장치 | |
| WO2024010248A1 (ko) | 딥러닝 기반 의료 데이터의 품질 개선 방법, 프로그램 및 장치 | |
| CN114723933B (zh) | 区域信息生成方法、装置、电子设备和计算机可读介质 | |
| WO2024014821A1 (ko) | 심전도 기반 신경망 모델의 학습 방법, 프로그램 및 장치 | |
| WO2023249411A1 (ko) | 데이터 처리 방법, 컴퓨터 프로그램 및 장치 | |
| CN111737575B (zh) | 内容分发方法、装置、可读介质及电子设备 |
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: 23827486 Country of ref document: EP Kind code of ref document: A1 |
|
| WWE | Wipo information: entry into national phase |
Ref document number: 18721804 Country of ref document: US |
|
| WWE | Wipo information: entry into national phase |
Ref document number: 2023827486 Country of ref document: EP |
|
| ENP | Entry into the national phase |
Ref document number: 2023827486 Country of ref document: EP Effective date: 20241108 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| WWP | Wipo information: published in national office |
Ref document number: 18721804 Country of ref document: US |