WO2012104779A1 - Traitement d'image, estimation de fréquence, commande mécanique et éclairage destiné à un système de commande et de surveillance tv automatique - Google Patents

Traitement d'image, estimation de fréquence, commande mécanique et éclairage destiné à un système de commande et de surveillance tv automatique Download PDF

Info

Publication number
WO2012104779A1
WO2012104779A1 PCT/IB2012/050434 IB2012050434W WO2012104779A1 WO 2012104779 A1 WO2012104779 A1 WO 2012104779A1 IB 2012050434 W IB2012050434 W IB 2012050434W WO 2012104779 A1 WO2012104779 A1 WO 2012104779A1
Authority
WO
WIPO (PCT)
Prior art keywords
light
drip
image
methods
light source
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
Application number
PCT/IB2012/050434
Other languages
English (en)
Inventor
Kai KAI TAO
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.)
Individual
Original Assignee
Individual
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
Priority claimed from US13/019,698 external-priority patent/US20120197185A1/en
Priority claimed from US13/356,632 external-priority patent/US20140327759A1/en
Application filed by Individual filed Critical Individual
Priority to CN201280016765.3A priority Critical patent/CN104010676A/zh
Publication of WO2012104779A1 publication Critical patent/WO2012104779A1/fr
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61M—DEVICES FOR INTRODUCING MEDIA INTO, OR ONTO, THE BODY; DEVICES FOR TRANSDUCING BODY MEDIA OR FOR TAKING MEDIA FROM THE BODY; DEVICES FOR PRODUCING OR ENDING SLEEP OR STUPOR
    • A61M5/00—Devices for bringing media into the body in a subcutaneous, intra-vascular or intramuscular way; Accessories therefor, e.g. filling or cleaning devices, arm-rests
    • A61M5/14—Infusion devices, e.g. infusing by gravity; Blood infusion; Accessories therefor
    • A61M5/1411—Drip chambers
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61M—DEVICES FOR INTRODUCING MEDIA INTO, OR ONTO, THE BODY; DEVICES FOR TRANSDUCING BODY MEDIA OR FOR TAKING MEDIA FROM THE BODY; DEVICES FOR PRODUCING OR ENDING SLEEP OR STUPOR
    • A61M5/00—Devices for bringing media into the body in a subcutaneous, intra-vascular or intramuscular way; Accessories therefor, e.g. filling or cleaning devices, arm-rests
    • A61M5/14—Infusion devices, e.g. infusing by gravity; Blood infusion; Accessories therefor
    • A61M5/168—Means for controlling media flow to the body or for metering media to the body, e.g. drip meters, counters ; Monitoring media flow to the body
    • A61M5/16886—Means for controlling media flow to the body or for metering media to the body, e.g. drip meters, counters ; Monitoring media flow to the body for measuring fluid flow rate, i.e. flowmeters
    • A61M5/1689—Drip counters
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61M—DEVICES FOR INTRODUCING MEDIA INTO, OR ONTO, THE BODY; DEVICES FOR TRANSDUCING BODY MEDIA OR FOR TAKING MEDIA FROM THE BODY; DEVICES FOR PRODUCING OR ENDING SLEEP OR STUPOR
    • A61M2205/00—General characteristics of the apparatus
    • A61M2205/33—Controlling, regulating or measuring
    • A61M2205/3306—Optical measuring means
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61M—DEVICES FOR INTRODUCING MEDIA INTO, OR ONTO, THE BODY; DEVICES FOR TRANSDUCING BODY MEDIA OR FOR TAKING MEDIA FROM THE BODY; DEVICES FOR PRODUCING OR ENDING SLEEP OR STUPOR
    • A61M2205/00—General characteristics of the apparatus
    • A61M2205/33—Controlling, regulating or measuring
    • A61M2205/3331—Pressure; Flow
    • A61M2205/3334—Measuring or controlling the flow rate

Definitions

  • TITLE Image Processing, Frequency Estimation, Mechanical Control and Illumination for an Automatic IV Monitoring and Controlling system
  • This invention relates to an IV monitoring and control system whose monitoring is done by video and image processing, whose dripping speed is measured using frequency estimation techniques, whose dripping rate is controlled using mechanical component and whose illumination is done with optical components.
  • infusion pumps are widely used as an automatic IV controlling device. Most infusion pumps does not monitor IV speed but controls the speed using mechanical devices, most commonly peristaltic pump.
  • Image processing techniques to process the video/image for extracting periodic measurement of the IV dripping process include:
  • Image enhancement techniques which further includes gray-level transformation, frequency-domain processing and wavelet techniques.
  • Thresholding techniques which further includes iterative method, arbitrary/constant or manually assigned/determined threshold level, and mean/median or other simple thresholding method.
  • Frequency estimation would estimate dripping frequency from periodic signal extracted from image sequence, disclosed methods include:
  • Non-parametric methods which further includes naive time-domain methods, time- domain statistical methods, Fourier and Fourier-related methods, and wavelet transform.
  • Parametric methods which further includes auto-regressive or auto-regressive mean- average spectrum estimation methods and eigenvector/subspace methods.
  • Apparatus controls the dripping speed by pressing the tube.
  • Apparatus include tube presser and supporter, use of leadscrew and differential leadscrew, use of lever, use of linear motion guide, rotational presser and cam embodiment.
  • Illumination system ensures the quality of captured video/image(s).
  • Principles, methods and apparatus include principles of reflection/brightness contrast reduction, multiple light sources, multiple sources from secondary light source, light source from mirror reflection, magnified light source from lens, use reflective surface of any level of smoothness, avoid shooting the reflection/brightness contrast and the use of light director/blocker.
  • Fig. 1.1-1 A show the image of drip chamber and Fig. 1.1 -IB shows within Fig. 1.1-lA an area used where image analysis is performed on.
  • Fig. 1.1 -2A shows a vertical Sobel gradient
  • Fig. 1.1 -2B shows a vertical Prewitt gradient
  • Fig. 1.1 -2B shows a Laplacian operator
  • Fig. 1.1-3 shows an image and its Sobel, Prewitt and Laplacian result.
  • Fig.1.1 -4A to Fig.1.1 -4D shows analysis steps performed on a sequence of captured images. Each figure contains on its top left the original image, top right the result of Sobel gradient operator, bottom left thresholding result of the Sobel gradient, bottom right erosion result of the bottom left.
  • Fig. 1.1-5 shows an erosion kernel used Fig. 1.1 -4A to D.
  • Fig. 1.1 -6A shows drip height from speed II ⁇ 13 periods dripping video.. ⁇
  • Fig. 1.1 -6B shows DFT of Fig. 1.1 -6A.
  • Fig. 1.1 -7A shows drip size from speed II ⁇ 13 periods dripping video.
  • Fig. 1.1 -7B shows DFT of Fig. 1.1 -7A.
  • Fig. 1.1 -8A shows region's average gray level from speed II ⁇ 13 periods dripping video.
  • Fig. 1.1 -8B shows DFT of Fig. 1.1 -8A.
  • Fig. 1.2.1-1 shows the comparison between image gradients, power-law and exponentiation transformation result.
  • Fig. 1.2.1-2 shows power-law result followed by Otsu thresholding and erosion.
  • Fig. 1.2.3-3 is the signal obtained by piece-wise transformation, followed by Otsu thresholding, erosion and maximum component. Drip size and height in upper and respective DFT in the lower.
  • Fig. 1.2.4-1 shows histogram equalization's effect thresholding algorithms
  • Fig. 1.2.4-2 shows histogram matching result for image enhancement.
  • Fig. 1.2.4-3 shows signal obtained by histogram matching, followed by Otsu thresholding, erosion and maximum component. Drip size and height in upper and respective DFT in the lower.
  • Fig. 1.3-1 shows how to perform frequency filtering that is equivalent to a spatial domain filter.
  • Fig. 1.3-2 shows how to convert Vertical Sobel mask to a convolution kernel
  • Fig. 1.3-3 shows frequency-domain high-pass filter effect.
  • Fig. 1.3-4 shows another example of frequency-domain filtering.
  • Fig. 1.4- 1 shows the result of wavelet filtering.
  • Fig. 1.4-2 shows the signals obtained by wavelet filtering followed by Otsu thresholding, erosion and maximum connected components and their DFT.
  • Fig. 1.5.1-1 compares Iterative method and Otsu's method
  • Fig. 1.5.1-2 shows the signals obtained by iterative method thresholding, preceded by Sobel gradient and followed by erosion and maximum connected components, together with their DFT.
  • Fig. 1.5.2-1 compares constant level threshold with Otsu and Iterative method.
  • Fig. 1.5.2-2 shows the signals obtained by constant level thresholding, preceded by Sobel gradient and followed by erosion and maximum connected components, together with their
  • Fig. 1.5.3-1 compares Otsu, mean and median thresholding.
  • Fig. 1.5.3-2 shows the signals obtained by mean thresholding, preceded by Sobel gradient and followed by erosion and maximum connected components, together with their DFT.
  • Fig. 1.5.3-3 shows the signals obtained by median thresholding, preceded by Sobel gradient and followed by erosion and maximum connected components, together with their DFT.
  • Fig . 2.2.1.1- -1 shows
  • Fig. 2.2.3.3-1 shows auto-correlation of drip height speed II signal, and correlogram, and DFT.
  • Fig. 2.2.3.4-1 shows DTFT of auto-covariance for drip height signal.
  • Fig. 2.2.3.4-2 shows DTFT of auto-covariance for drip size signal.
  • Fig . 2.2.3.5- -3 shows Fig. 2.2.3.5-4 shows incorrect and correct ways DST-II extension.
  • Fig. 2.2.3.5-5 shows magnitude of DST-II coefficients for drip height signal.
  • Fig. 2.2.3.5-6 shows magnitude of DST-II coefficients for drip size signal.
  • Fig. 2.2.4-1 shows wavelet period counting for speed I drip height signal.
  • Fig. 2.2.4-2 shows wavelet period counting for speed II drip height signal.
  • Fig. 2.2.4-3 shows wavelet period counting for speed III drip height signal.
  • Fig. 2.3.1.1-1 shows Yule- Walker method for speed I drip height
  • Fig. 2.3.1.1-2 shows Yule- Walker method for speed II drip height
  • Fig. 2.3.1.1-3 shows Yule- Walker method for speed III drip height
  • Fig. 2.3.2.1-1 shows MUSIC method pseudospectrum for speed I drip height signal
  • Fig. 2.3.2.1-2 shows MUSIC method pseudospectrum for speed II drip height signal
  • Fig. 2.3.2.1-3 shows MUSIC method pseudospectrum for speed III drip height signal
  • Fig. 2.3.2.1-4 shows MUSIC method pseudospectrum for speed I drip size signal
  • Fig. 2.3.2.1-5 shows MUSIC method pseudospectrum for speed II drip size signal
  • Fig. 2.3.2.1-6 shows MUSIC method pseudospectrum for speed III drip size signal
  • Fig. 3.1-1 shows IV speed adjuster used for manual adjustment
  • Fig.3.1-2 shows side or front view for possible shapes of IV tube presser/supporter
  • Fig.3.1-3 shows axial/top/bottom view for possible shapes of IV tube presser/supporter
  • Fig.3.1-4 shows shape, edge, angle and ways of contact between IV tube, presser and supporter.
  • Fig. 3.1-5 shows a concrete example of a presser with sharp edge in its top, left, side and right view.
  • Fig. 3.2-1 shows side and axial view of a leadscrew
  • Fig. 3.2.1-1 shows a differential leadscrew combination is shown where the precision enhanced 10 times. Dimensions are purely illustrational.
  • 3.3-1 shows off-axis movement illustration
  • Fig. 3.3-2 shows key/keyway combination to control off-axis movement
  • Fig. 3.3-3 shows spline/groove combination to control off-axis movement
  • Fig. 3.3-4 shows bearing(s) to control off-axis movement
  • Fig. 3.4-1 shows use of lever in translating motion
  • Fig. 3.5.1-1 shows the pivoted "nutcracker"
  • Fig. 3.5.1-2 shows the principle of off-axis movement can be absorbed by the pivoted "nutcracker.
  • Fig. 3.5.1-3 shows the leverage of the pivoted "nutcracker.
  • Fig. 3.5.1-4 shows the linearly moving part might contact the rotational part at any location, in any geometric configuration.
  • Fig. 3.5.2-1 shows a rotational pivoted "Nutcracker"
  • Fig. 3.6-1 shows the use of cam
  • Fig. 4.1-1 shows example of good illumination
  • Fig. 4.1-2 shows example of bad illumination
  • Fig. 4.1-3 shows the cause of reflection/brightness contrast
  • Fig. 4.2-1 shows by increasing the distance between light source and drip chamber reflection/brightness contrast might be reduced.
  • Fig. 4.2-2 shows by mutual cancellation of brightness unevenness of multiple light sources reflection/brightness contrast might be reduced.
  • Fig. 4.3-1 how multiple light sources can be used.
  • Fig. 4.4-1 shows how a single light source might be directed by light guide/light tube/light pipe/integrator bar/optical fiber to illuminate drip chamber from multiple locations.
  • Fig. 4.4-2 shows the principle of light guide/light tube/light pipe/integrator bar/optical fiber in creating multiplicity of images for a single point source.
  • Fig. 4.5-1 shows mirror of mirror combination might be used to direct light
  • Fig. 4.6-1 shows how light source might be magnified to cancel unevenness of individual point sources.
  • Fig. 4.7-1 shows how reflective surfaces might be used to reduce reflection/brightness contrast.
  • Fig. 4.7-2 shows the reflective surface can take different ways of formation and shapes.
  • Fig. 4.8-1 shows how a rough surface might be used to cause light to scatter randomly.
  • Fig. 4.10-1 shows light director/blocker extending from light source to object
  • Fig. 4.10-2 shows light director/blocker and image capturing can be arranged in any relative position as long as reflection/brightness contrast in the view of the image capturing device can be reduced.
  • Fig. 4.10-3 shows light director/blocker can be put in different places.
  • FIG. 0.1- 1 A schematic for the whole system is shown in Fig 0.1- 1 which comprises subsystem of illumination, subsystem of image capturing, processing and frequency estimation and subsystem of mechanical control.
  • ⁇ method A, method B ⁇ x ⁇ data I, data II ⁇ would mean applying each of the two methods on each of the two datasets, which can also be denoted by ⁇ (method A, data I), (method A, data II), (method B, data I), (method B, data II) ⁇ .
  • Fig. 1.1-1 A shows an image of an IV drip chamber
  • Fig. 1.1 -IB we use a rectangle close to the dripping mouth to specify the area where the image processing will be taken on.
  • the purpose of choosing an area close to the drip mouth is primarily to enable processing with low resolution, low frame -rate image capturing device and low speed processor. Real implementation could monitor any area if periodicity signal can be extracted.
  • Fig. 1.1 -2A shows a vertical Sobel gradient operator
  • Fig. 1.1 -2B shows a vertical Prewitt gradient operator
  • Fig. 1.1 -2C shows a Laplacian operator.
  • Fig. 1.1 -4A to D show image processing steps for four images in a sequence.
  • the upper left is the original image with the index in the sequence shown in the title
  • the upper right we show the example of applying Sobel gradient (vertical + horizontal).
  • the lower left applies Otsu's thresholding method on Sobel gradient result and converted a gray level image to a binary image.
  • the first number shows the number of connected components
  • the second number shows the size of the maximum connected component measured in the number of pixels
  • the third number shows the average height of the y coordinate of the maximum connected component with y increases from top to bottom in the image.
  • DFT both recognized the correct number of periods by the index of the non-DC component with the maximum DFT magnitude.
  • Fig. 1.1 -8A we show for the same image sequence a signal extracted by very crude, most simplistic and very improbable a means for extracting a meaningful signal: simply taking the average of all pixels' gray level value. Not a single image processing technique has been applied, yet the signal Fig. 1.1 -8A still shows regular periodic pattern and its DFT in Fig. 1.1- 8B also recognizes the correct period count.
  • PPlleeaassee rreeffeerr ttoo [[CChh..33,, GGoonnzzaalleezz,, RR..CC..,, WWooooddss,, RR..EE..,, DDiiggiittaall IImmaaggee PPrroocceesssiinngg,, 22eedd,, PPrreennttiiccee HHaallll,, 22000022]] ffoorr ddiissccuussssiioonn..
  • ⁇ is an offset added to x .
  • x would usually be normalized to [ ⁇ , l] , and ideally c ( + ⁇ ) ⁇ would also map [ ⁇ , l] to [ ⁇ , l] , but in practice this does not need to be strictly observed.
  • Fig. 1.2.1-1 The left-up corner shows an unprocessed image (21 means it is the 21 st image in a sequence). Sobel, Prewitt and Laplacian gradient results are shown in the
  • Fig. 1.2.1-2 shows the ensemble result of power- law transformation with thresholding, erosion discussed in application US 12804163.
  • the lower-left image show thresholding result, and the lower-right image is the result after erosion so that smaller parts are removed. From the lower-right image, the vertical location and height of the drip is extracted by finding the maximum connected components in the image, and would be stored into a vector. We could clearly see in the processing steps how the extracted information matches well with our visual interpretation.
  • Fig. 1.2.1-3 shows how dripping speed is measured from the drip height.
  • Fig. 1.2.1-4 shows how dripping speed is measured from the drip size.
  • exponentiation transformation which is also called inverse-log transformation since log and exponentiation are the inverse.
  • a is positive and usually greater than one.
  • c is also usually a positive number.
  • x would usually normalized to [ ⁇ , l] , and ideally c ⁇ - would also map [ ⁇ , l]to [ ⁇ , l] ,
  • Fig. 1.2.2-3 upper part is the drip size data over 180 samples, and the lower its DFT result. Same period count as drip height signal is given by DFT.
  • gray-level y 5 ⁇ (3 X — 2) give correct periodic drip height signal when works in conjunction with Otsu threshold, erosion and maximum connected components for all speeds I, II and III. From the resultant drip height signal DFT gives the correct period count.
  • Piecewise-linear function transformation as described in [ ⁇ 3.2.4, Gonzalez, R.C., Woods, R.E., - Digital Image Processing (2ed)], is a method complementary to other gray-level transformation techniques with the advantage that it can approximate arbitrarily complex function. All previous can be mimicked by using it. In hardware implementation, it is equivalent to what is normally called "look-up table".
  • histogram information can be used for gray-level transformation. This class of techniques are commonly used and a detailed description can be found in [ ⁇ 3.3, Gonzalez, R.C., Woods, R.E., - Digital Image Processing, 2ed, Prentice Hall, 2002].
  • discrete intensity values of a gray-level image may span an interval [0, L— 1] and the probability for each discrete intensity value is n N, where N is the total number of pixels and 7ij the number of pixels having intensity i.
  • the span [0, L— 1] might be normalized to [0, 1] and we denote the random variable corresponding to the normalized interval as r.
  • the probability density function is continuous and strictly monotonically increasing so that its inverse always exists.
  • the probability density function is denoted as
  • histogram matching The process of matching/specification the histogram of an image to an arbitrarily assigned histogram is called histogram matching or histogram specification.
  • histogram equalization When the target histogram is simply of a uniform distribution, it is called histogram equalization.
  • histogram equalization can be subsumed as a special case under histogram matching, and we do not rule out the possibility of using it.
  • Fig. 1.2.4-2 shows histogram matching for several images in which the target is the Sobel gradient (horizontal + vertical) result of image 21 in speed II « 13 sequence. Decent results have been achieved and we could also see how close the histograms of the result match the target.
  • histogram matching/specification including equalization
  • Gray-level transformation as a general class of techniques, can be used in the processing steps of a video/image processing based IV monitoring system. No other techniques fall in this class could have any substantial difference with our disclosed methods.
  • Frequency-Domain Processing Frequency Domain techniques are frequently used in image processing. They work by first transforming the image into its frequency domain representation, apply processing techniques and inverse-transform the result into spatial domain.
  • an digital image as 3 ⁇ 4) , 0 ⁇ a ⁇ - 1, 0 ⁇ b ⁇ N - I , and make periodic extension to coordinates outside [O, M - 1] x [O, N - 1] so that
  • Periodic extension enables us to define periodic convolution
  • the above theorem means that the effect of each spatial domain filter using the periodic extension definition can be achieved via frequency domain multiplication.
  • Fig.1.3-1 The process of doing Sobel vertical filtering in frequency is shown in Fig.1.3-1: 1. Take DFT of the original image. Shown in Fig. 1.3-1 lower left is the shifted version which moved coordinate [ ⁇ , 0] to center, which is a convention in the field.
  • Spatial domain filters are typically arranged in (2 + l) x (2 + l) matrix and the weight corresponding to each image pixel itself is at the center (k + 1, k + 1) .
  • spatial domain filtering it will be directly "masked" on the original image, multiply with underlying pixel and take the sum;
  • convolution since only convolution has a direct frequency-domain counterpart but not masking, we need to change it to equivalent convolution first:
  • the 2 nd image in the 1 st row shows logarithmic scale 2D DFT of the original image.
  • the filtered and shifted spectrum is shown in the 3 rd image, which clearly has a dark/empty center due t the filter effect.
  • the reconstructed image is shown in the 4 th image but rather dark and the scaled display is in 5 th image of the 1 st row, showing how contrast has been enhanced.
  • Spatial domain filtering can be achieved in frequency domain.
  • Filters designed purely in the frequency domain can also be used to enhance the image.
  • c horizontal is essentially a vertical gradient taking at two upper and two lower pixels together
  • c vertical is essentially a horizontal gradient taking two left and two right pixels together
  • a tick / means the wavelet gradient result followed by Otsu thresholding, erosion and maximum component is accurate in that DFT gives the correct period count.
  • Thresholding is one of the most basic image processing techniques. In application US 12804163 we show that Otsu's method can be used to automatically detect threshold level. In this disclosure we are going to show other methods also work.
  • Iterative method finds a thresholding level L in an iterative process. Its implementation is simple, requiring no specific knowledge of the image and is robust against noise.
  • Otsu's method is representative of the class of thresholding algorithms that uses histogram information; with iterative method, we have shown that automatic thresholding can also be done without explicitly using histogram information.
  • Fig. 1.5.3-1 mean and median of image pixels as thresholding value are compared with Otsu threshold in each quadrant, also followed by erosion and maximum connected component. Although visually they do no convey very good information on the size and location of the drip, the final signal extracted after erosion and maximum connected component, have been found:
  • frequency estimation has many synonyms. “Spectral/spectrum” can be used in place of “frequency”, and “analysis/detection” are also used in many occasions instead of “estimation”. Terms like “period/periodicity” as well as “count” are also commonly used. It is believed that the choice of terms, if appears to be different from I am using in this disclosure, would not render the claims of this application inapplicable since it is the underlying methods that precisely defining the scope of protection, rather than the particular choice made on the naming of the methods.
  • DFT Discrete Fourier Transform
  • MUSIC Multiple Emitter Location and Signal Parameter Estimation
  • QualityO assigns a numerical value with higher one represents a higher signal quality.
  • any periodic measurement can be used. It is self-evident that drip height is a better periodic signal comparing with region's average gray level, but the fact that region's average gray level might not pass the test of some of the algorithms listed below is primarily due to the way we were extracting the signal.
  • video/image processing is done in a very small video window near the drip chamber's mouth where drips are forming and starting to fall.
  • the window size is smaller than 20(width)x50(height), fewer than 1000 pixels.
  • ⁇ ( ⁇ ) can be a constant smaller number independent of N, such as 0.5 or 1 ; Or it can be a monotonically increasing function of N. Intuitively, if you have only a signal of length 10, an error of 1 period is of course intolerable; but if you have a signal of length 1000, the same error would certainly be within the margin.
  • Suitability (S m y example signal' E)— 0 SuitabiUty( y 0ur improved signal ⁇ E)— 0 And what really determines is the quality of S as well as the ⁇ ( ⁇ ) one chooses.
  • the threshold of 13 was not automatically computed, but manually picked.
  • end return Index _ record is the average distance between successive indices of "crossing" points, and if two successive indices is closer than ⁇ ⁇ mean it means one of them might correspond to a small spike or a splitting peak, ⁇ as 0.5 has been tested as an appropriate value for the drip height signal of speed I, II and III.
  • R xx (S, m) is symmetric about 0
  • period T and k is an integer.
  • a monitoring and controlling device that is capable of only recognizing integer periods cannot determine whether the speed has reached 6.2 drips or not after monitoring for a 6- second period but can only know, after repeated adjusting and monitoring, that the speed is now in the range of [6, 7] . To approximate the speed as close as possible, it has to extend the observing period to much longer, in this case at least 30 seconds, since 62 has only two divisors smaller than itself: 2 and 31. Even if at 30 seconds it observed a drip count of 31 drips, the actual speed would still be between
  • Converge is a mathematical term and its use here is to mean that after repeated adjust- monitor feedback loop, the actual speed of drip finally falls into the tolerance range of the prescribed value.
  • Fig. 2.2.2.1-3 and Fig. 2.2.2.1-4 show auto-correlation also works for drip size signal for speed I, II and III.
  • the biased auto-covariance of a real sequence S is defined as
  • V xx (S, m) V furnish(S -m) it is therefore suffice only to compute for positive m 's.
  • the determination of periods is done by counting the distance between V XX (S, 0) and the next local maxima. Like ior R xx (S,m) , V ⁇ S ⁇ T ⁇ would attain local maxima, in which T E denote the estimate of period T and k is an integer.
  • auto-covariance is also one of the recommended methods in actual implementation.
  • AMDF Average Magnitude Differential Function
  • the Biased Average Magnitude Differential Function (AMDF) of a real sequence S is defined as in which we use an overline on top of D for differentiation with the unbiased version.
  • AMDF Biased Average Magnitude Differential Function
  • AMDF Unbiased Average Magnitude Differential Function
  • k can be chosen to be any positive value, but integers like 1 , 2 are commonly used. And a property is that
  • Fourier transform has a vast number of variations and derivatives and it's impossible to exhaust.
  • a clear distinction between Fourier- family methods and previous methods is that it whereas previous methods uses time domain signal directly, Fourier methods would estimates its constituent components at different frequencies, e 1' can take either discrete or continuous 6 .
  • DTFT discrete -time Fourier transform
  • DTFT discrete -time Fourier transform
  • Periodogram and DTFT are related by
  • periodogram is simply the square of DTFT' s magnitude divided by N . Both of them can be used to estimate signal frequency for our application. And advantage of periodogram and DTFT over DFT is that they estimate fractional frequency. This is also achievable by auto-correlation, auto-covariance and AMDF as well as many other algorithms described afterwards, but with different principle. In clinical application this would result in quick convergence speed which is an important improvement over DFT speed counting. Please refer to [ ⁇ 2.2.2.1 auto-correlation] for discussion.
  • D((o) and ⁇ ) at these ⁇ 1 ' s could be larger than D co) and ⁇ ) at higher co s which corresponding to the AC components of the signal, and would therefore cause problem if we compare D(co) or P(co) magnitude to estimate the signal frequency.
  • the simplest solution is to remove the signal's mean.
  • the frequency estimation would be done by simply scanning the DTFT/periodogram sequence.
  • the DTFT/periodogram peak for drip height signal of speed II « 13 periods is at
  • the Bartlett's periodogram average is one of the variations of periodogram and is defined as
  • m which is used to finely divide Aco increments in Fig. 2.2.3.2-1 is also chosen to be 100, and real implementation can use any value.
  • periodogram and DTFT for efficient computation and more accurate estimation one could first use DFT to locate the interval and then compute Bartlett's periodogram on the vicinity of the interval.
  • the correlogram of a real sequence S is defined as in which R xx (S, k) is the auto-correlation sequence as defined in [ ⁇ 2.2.2.1 Auto-correlation] .
  • correlogram result is found to be very close to them. This accuracy of correlogram is therefore confirmed.
  • the auto-covariance V xx sequence can be either biased or unbiased.
  • DCT and DST are variations of the DFT representing real sequence S with real coefficients. Depending on the different choices of defining periodic and symmetric extension, there are at least 16 different variations of DCT and DST. Please refer to [Wang, Z. 1984., Fast
  • the first N DCT-II coefficients already contain the full information of the extended sequence, and S DCT _ n can be reconstructed by first multiplying DCT - II sequence with respective e 2 and then take the inverse DFT.
  • the first N result of the inverse DFT would be the original sequence S .
  • Fig. 2.2.3.5-l show the DCT-II extension for drip height, speed II ⁇ 5.3 periods signal compared with the original. Note how the symmetric extension added ambiguities to the signal in terms of periodicity.
  • Fig. 2.2.3.5-2 still show correct integer level precision period count for drip height signal of speed I, II and III.
  • DST-II differs with DCT-II only in that it not only "flip” the original sequence, but also invert (take negative) them so that odd symmetry new sequence cancels cosine coefficients while preserving sine sequence.
  • the correct way of extension is to first subtracting mean from the original sequence then do the normal DST-II extension.
  • wavelet transform The basic ideas behind wavelet transform are multi-resolution and filter banks. There is a low pass filter and a high pass filter. After a signal is passed through the low-pass filter and down- sampled (denoted byj), the result represents the low-frequency local component of the signal; on the other hand, the down-sampled high-pass filter result represents the high-frequency local component of the signal.
  • the low- frequency local component is simply the local average of adjacent components
  • the high-frequency local component is the difference between two adjacent components.
  • the IV dripping process can largely be assumed to be a stationary type of signal expect during moments when the patient is moving his arm or due to other activities.
  • wavelet transform can still be used to detect periods for this type of signal, and the principle is to some extent similar with time-domain methods.
  • Results are shown in Fig. 2.2.4-1 to Fig. 2.2.4-3.
  • the algorithm with parameters above recognizes correctly peaks at all speeds I, II and III.
  • the peak locations are marked with upward-pointing triangles.
  • the 3 rd and 4 th level of approximate reconstruction are also shown below the 2 nd approximate reconstruction from which we can see that excess levels of approximation might flatten the signal too much to cause peak detection fail.
  • the Daubechies family of wavelet can have different lengths, and there are many other types of wavelets such as ⁇ biorthogonal, cubic spline, Haar, Mexican hat, Morlet, Meyer, symlets ⁇ and customarily constructed types. But none of them would constitute substantial difference from our algorithm.
  • Daubechies D2 wavelet is equivalent to Haar wavelet, and wavelet transform with it is also called Haar transform. Depending on the quality of the signal, it is reasonable to expect that others types of wavelets and parameters can be used for peak (peak) counting in this application.
  • the parametric methods assume that the signal satisfies a generating model with known functional form and then proceed by estimating the parameters in the assumed model. The signal's spectral characteristics of interest are then derived from the estimated model.
  • Yule- Walker is a method for estimating frequency for AR models. For a real sequence S its steps are:
  • ⁇ ⁇ k — ⁇ x[n] x[n + k]
  • biased auto-correlation is used to ensure the matrix is positive definite.
  • Unbiased auto-correlation can also be used and please refer to [Hayes, M. - Statistical Digital Signal Processing and Modeling (Wiley, 1996)] for detail.
  • S in general is a periodic signal, and obviously depends on its previous values. AR model fits this physical basis best; poor estimation might occur from MA model since its assumption is not consistent with the reality.
  • MUSIC Multiple Emitter Location and Signal Parameter Estimation
  • pseudospectrum or eigenspectrum
  • Fig. 2.3.2.1-1 to Fig. 2.3.2.1-6 show these methods work for ⁇ drip height, drip size ⁇ x ⁇ speed I, II and III ⁇ .
  • drip size signal larger P such as over 80 might be needed, depending on the quality of the signal.
  • Fig. 2.3.2.1-1 to Fig. 2.3.2.1-6 are shown in logarithmic scale. There are two reasons: (1) if it shown in linear scale, many of the smaller values would be hardly visible. (2) what is displayed is actually pseudospectrum, so we don't have to make them as in linear scale as shown for previous experiments.
  • frequency can also be estimated not by via any means to estimate the original power spectrum (periodogram or any other), but via its pseudospectrum through eigenvalue decomposition.
  • pseudospectrum methods can also be used for our IV speed monitoring application.
  • Fig.3-1 is a general schematic of the mechanical sub-system of our IV monitoring and control system. A leadscrew is shown more prominently than other parts because we believe it is essential to the system. However, we stress that no limitation is made here that the real implementation must use leadscrew and its function can be substituted by other parts.
  • Fig. 3.1- 1 Devices such as Fig. 3.1- 1 are used in conventional gravity based dripping for adjusting drip speed. It works by pressing the tube to adjust its thickness. The presser rolls in a groove and changes the thickness of the tube by pressing it.
  • Fig.3.1-2 shows the side (or front) view of some possible shapes of back supporter for tubes.
  • Fig.3.1-3 Viewing from the perspective of tube's axial direction (top/bottom), there can also be a numerous matching and complementary shapes to form the presser/supporter combination. Five examples are shown in Fig.3.1-3.
  • Fig.3.1-4 On the left of Fig.3.1-4, from top to bottom, we give example that the shape of the contacting point between presser and tube can of an angle, or flat, or rounded.
  • angle can be obtuse, right or acute; it can either taper or expand; it can either have a sharp edge or a flat surface at the contacting point.
  • presser and supporter can cause the flow in the tube to stop, then they the two could be used a presser/supporter pair.
  • the presser is drawn to have inner thread so it could be used as a nut to be mounted with a leadscrew so that leadscrew's rotation would be translated into presser's axial movement. It is also possible that the presser is connected with a linearly moving part directly without requiring leadscrew-nut combination. In both cases there could be off-axis movement and it could be reduced with techniques in [ ⁇ 3.3 Linear Motion Guide].
  • presser/supporter pairs shown above are used for a linear actuator is pushing the presser directly toward the direction of the tube. Besides linear actuator, there are still
  • Fig.3-1 Please refer to Fig.3-1 for the general schematics of the mechanical subsystem.
  • a mechanism for IV control has the following characteristics: 1. High resolution so that the control can be accurate.
  • leadscrew is an ideal solution which satisfies 1-3.
  • Leadscrew is a basic mechanical structure and is known to everyone work in mechanical engineering. Please refer to [ ⁇ 8.2, Shigley's Mechanical Engineering Design] for discussion and properties.
  • d mean diameter of the ring on which screw and nut touch
  • Property I High resolution so that the control can be accurate.
  • Lead / is usually very small.
  • the linear stepper motor, / could be made as small as 0.5mm, and each step's rotation could be made to as small as 7.5 ° so that the stroke of each step is only
  • PPrrooppeerrttyyy IIII SSeellff-- lloocckkiinngg ssoo tthhaatt nnoo eenneerrggyy iiss rreeqquuiirreedd ffoorr mmaaiinnttaaiinniinngg ccoonnttrrooll ppoossiittiioonn..
  • Property III Strong output force so that sufficient pressure can be applied on IV tube.
  • leadscrew is an ideal mechanical structure for controlling IV dripping speed.
  • a linear motor, or any motor is an assembly and combination of different
  • Leadscrews can be used in many other places besides within linear motor. In fact, as Fig.3- 1 would show, it might appear for one or more times at different places for translating rotary motion into linear. For example, a linear rail/slide could have leadscrew, but it might not use the electrical part of motor at all so that naming it as a "motor” is obviously inappropriate. It is therefore better to use the name "leadscrew” alone to direct our attention to its distinct properties.
  • the pitch on the presser and its corresponding shaft part to be 1.0mm and on the tube fixture and its corresponding shaft part to be 0.9mm.
  • there of course needs to be bearings, key/keyway or spline/groove combination to prevent them from rotating, which are omitted in the image for visual clarity.
  • the presser needs to be driven 30mm to the right whereas the supporter with the tube are driven 27mm to the right, and their relative movement results in the full 3mm pressing of the tube.
  • key/keyway or spline/groove is on the inside or outside.
  • leadscrew as the linear actuator in the two figures, we by no means require that spline/groove or key/keyway must be used on leadscrew driven parts. They can be used to guide linear motion resulted from any component(s).
  • bearings can be either on the outside or fit within groove/channel/track cut within the moving parts.
  • Fig. 3.4-1 shows one of the numerous possible ways of creating a linear motion in that a leadscrew first causes the rotation of one side of a lever, the rotation of the other end of the lever then causes the linear movement of a slider/presser.
  • the nut of the leadscrew has a small cylindrical connector that is fitted into the groove in the lever and the slider/presser also has a same connector fitted into the groove in the lever. Bearings have been used for both the leadscrew nut and the slider but of course spline/groove and key/keyway can also be used.
  • lever length ratio shown here are only for illustrational purposes. Levers can be classified into three classes according to the relative position of the fulcrum, the load and the force and the type that the load is between the force and the fulcrum can also be used.
  • This mechanism is simple, low cost, and has been tested to be very effective in operation.
  • the principle used by key/keyway, spline/groove and bearing is to prevent off-axis displacement by hold, grip or push firmly against it.
  • the pivoted nutcracker works by decomposing it.
  • Fig. 3.5.1-l shows the drawing.
  • the tube's supporter and presser are assembled together at a pivot on which the presser or both the presser and the supporter could turn about.
  • the fit at the connection should not be too tight to prevent the turning.
  • a linear motion either guided by ⁇ key/keyway, s line/groove, bearing, etc. ⁇ or not, now causes the presser arm to rotate about the pivot.
  • the tube back supporter would be fixed and not allowed to move, or it might also be allowed to rotate and be connected to a driving part.
  • the closing of the angle between the presser and tube supporter compresses the tube, and the opening does the reverse.
  • a groove of uniform width is cut in the presser.
  • the width of the groove is slightly larger than the diameter of a sphere or cylinder mounted at the head of the linear motion part so that when the linear parts moves forward or backward along its axis, the sphere or cylinder could have a relative sliding movement within the groove.
  • the length of the presser arm is usually much larger than then 3mm-around diameter of the tube, and the sphere/cy Under mounted at the linear motion part touches the groove sides also at a distance from the pivot usually much larger than 3mm tube diameter.
  • the opening angle between presser arm and the tube supporter, denoted by ⁇ would be an angle at most times smaller than 15 ° .
  • the pivoted arm In addition to solving the problem of off-axis displacement, the pivoted arm also enhances the precision by ratio of the horizontal distance D between sphere/cylinder's touching point to the pivot and the horizontal distance d between presser arm and tube's touching point to the pivot. By the principle of lever this also magnifies the force by D/d. Please see Fig. 3.5.1-3 in which this relationship is shown using similar triangle property. Recall that in [ ⁇ 3.3 Linear Motion Guide], we mentioned 0.0254mm (O.OOlinch) is a common stroke distance for linear stepper motors and it is roughly 1/1 18 of a 3mm inner diameter of an IV tube. With the use of this pivoted "nutcracker" presser, this 1 18 parts division could be further divided by a ratio of D/d which results in even finer precision over the dripping rate control.
  • the linear actuator pushes/drags the presser; it can push/drag either the presser arm or the tube supporter (when it allowed to move) to create relative movement; there can also be more than one such linear actuators to drive both of them.
  • the presser and tube supporter are connected at one end which allows the relative rotational movement between the two parts.
  • One or more linear actuator(s) that push(es) or drag(s) the presser and/or tube
  • the connector between the linear motion part and the arm groove would usually have a sphere or cylindrical shape.
  • a bearing is also allowed.
  • the essential characteristics is that is must have in at least one of the numerous cross-sectional surfaces a circular circumference which would allow it to move smoothly with in the groove. Other variations are possible at the cost of probably additional difficulty.
  • linearly moving part can contact the presser at any location in any geometric configuration, not necessarily on approximately the half-line/ray which start from the pivot and pass through the contacting point of the tube with the presser or supporter.
  • An example is shown in Fig. 3.5.1-4.
  • FIG. 3.5.2-1 shows a variation which is by connecting the shaft of a rotational motor to either of the presser or supporter so that the rotation of the motor could result in the change of angle between the presser arm and supporter. In this configuration sphere/cylindrical shape and groove is not needed.
  • presser and supporter are both driven by rotational components, or even in a mixed combination that one part's rotation is driven by a linearly moving part, and another part's rotation is imparted by a rotational component.
  • the essential result would always be the relative rotation of the two parts.
  • the presser and tube supporter are connected at one end which allows the relative rotational movement between the two parts.
  • Rotational actuator(s) connected to presser and/or tube supporter, directly or via other intermediate mechanism, whose rotation(s) result(s) in opening or closing the angle between them.
  • Cams can also be used to convert a rotary motion into linear. There are numerous types of cams and we have shown in this illustrational embodiment a spiral. There are five positions shown in Fig. 3.6- 1 to show how rotation of the cam would drive the linear motion of the presser.
  • the central circle in the front view is the motor shaft.
  • a board is connected to the shaft and a groove is cut on the board.
  • the geometric shape of the groove is the envelope a circle or any shape assuming circular circumference in at least one of its numerous cross-sections, running with its center moved along a spiral curve. If the groove rotates in the clockwise direction, the presser will be pressed to the right through the cylindrical connector that is fitted into the groove, and will be pulled to the left if the groove rotates in the anti-clockwise direction.
  • Cams are not self-locking so that a steady current might be needed to maintain its position. To lock the cam without continuous current, one might, in the following steps:
  • the plate has slots or holes evenly or unevenly spanned in different directions.
  • a cam in conjunction with bearing to guide the linear motion of the presser
  • a cam structure can also be used alone so that is edge directly touches the IV tube.
  • the change of ⁇ results in the change of radial length hence can be used to directly press the tube. It is also a possible implementation although issues such as slipping of the tube needs to be properly addressed.
  • cam can also be used in other parts of the system without directly driving the presser. It can be used to either translate rotary motion into linear or linear motion into rotary in any parts of the system, which is what it by its nature does. 4. Illumination
  • Fig. 4.1-l shows examples good illumination. LED lights in this example are projected from the top of the chamber through a light director/b locker shown in Fig. 4.10-1. The overhead lighting creates no spurious brighter spots from drip chamber surface's reflection in the image.
  • Fig. 4.1-2 show three examples of poor illumination where reflections of LED light on the drip chamber surface make video/image processing difficult.
  • the LED are placed on the side so that the dripping mouth and drip, although surrounded in brighter reflections of the chamber, are still visible. If light is projected from front or back of the chamber, the drip mouth and drip could be completely masked out by the brighter area(s) either due to light itself (from back) or its reflection (light in front).
  • the light power at dx is inversely proportional to the square of its distance from the point source:
  • the reflection interferes with drip detection for two reasons:
  • the area(s) caused by these reflections are usually brighter than reflections from the drip, or overlapped with the drip.
  • Primary light source A light source where a physical quantity of other form is converted into light. This includes LED light, incandescent light, infrared lights, ultraviolet lights, laser and any other type of light sources.
  • Secondary light source A light source whose lights are directed from one or more primary light sources by optical devices. Illumination using optical fiber, light-tube/light- pipe/integrator bar, assembly of mirrors' reflection, reflective surface all belong to this type.
  • the type of light source here can be either primary or secondary type.
  • the symbol of light source in Fig. 4.3- 1 looks like suggesting primary type, which is only for better illustration when comparing with Fig. 4.4- 1.
  • each of the multiple light sources there is also no requirement on how far should each of the multiple light sources be separated.
  • even a second light source as close to the drip chamber as the original light source could effectively cancel a large portion of unevenness in brightness. Therefore, it is perfectly possible to use an array of light sources close to, concentrated or near the original single light source location to cancel the unevenness of each.
  • Such array of light sources can also be manufactured integrated in a package, either LED or other type, which contains an array of light-emitting elements. This should also be considered as an instance of the multiple light sources.
  • Fig. 4.4-2 illustrates the principle of light tubes.
  • Each light source (whatever type) is in fact composed of numerous point sources.
  • the effect looks like numerous rays are coming from the point source's numerous virtual images so that it no longer behaves like a point source, but in effect similar to a scattered/diffusive light source. Please refer to [p.105, Smith, Warren J., Practical Optical System Layout and Use of Stock Lenses].
  • the principle of light tube is unique in itself. Although similar to the 2 nd "cancellation" principle of reflection/brightness contrast reduction, it is more appropriate to leave it as a single class alone rather than classified into the cancellation principle.
  • Bundles of the above Fig. 4.4-1 show that how a single light source (of any type) can be used to create multiple secondary light sources via light-tube. For thinner types of light-tube such as optical fiber, bundles of them can be used together.
  • Fig. 4.5-1 shows that assemblies of mirror can also be used to direct light so that multiple light sources can be created from a single one. It can of course be used just to direct a single light without creating a multiplicity of them.
  • Any shape of mirrors can be used since it is only used to direct light, not the image.
  • the shape might be arbitrarily curved or assuming particular geometric shapes. It also does not matter whether such mirror of shape, when used in ordinary occasions, might create some "bizarre" effect or not. Any type of mirror could be used as long as it could direct light.
  • each surface can be either convex, concave, or flat; for thick lens the ability of converging light also depends on its thickness; for combinations of lens the possibilities are impossible to enumerate.
  • EFL effective focal length
  • Parabola or paraboloid light rays emitting from the focus would be reflected so that all are become parallel with the axis (the symmetric axis of the shape itself).
  • Hyperboloid or hyperbola light ray emitting from one focus would have their
  • backward extension lines converge at the other focus so that it looks like the light is coming from the virtual image at the other source.
  • a light blocker (which can either be made/integrated as part of the light source to prevent it from scattering light to all directions, or separate from the light source) that blocks the direct path between the light source and the drip chamber.
  • a reflective surface whose reflection the light source uses to illuminate the object indirectly.
  • a rough surface is shown in Fig. 4.8-1 and it also requires a light blocker to prevent light from illuminating the drip chamber directly.
  • a light blocker (which can either be made/integrated as part of the light source to prevent it from scattering light to all directions, or separate from the light source) that blocks the direct path between the light source and the drip chamber.

Landscapes

  • Health & Medical Sciences (AREA)
  • Vascular Medicine (AREA)
  • Engineering & Computer Science (AREA)
  • Anesthesiology (AREA)
  • Biomedical Technology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Hematology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Animal Behavior & Ethology (AREA)
  • General Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • Veterinary Medicine (AREA)
  • Physics & Mathematics (AREA)
  • Fluid Mechanics (AREA)
  • Image Analysis (AREA)
  • Image Processing (AREA)

Abstract

L'invention concerne un dispositif, un appareil mécanique et un système d'éclairage destinés à un système de surveillance d'intraveineuse (IV). Le dispositif extrait un signal périodique du procédé d'égouttement IV par utilisation de techniques de traitement d'image ou de vidéo et mesure la vitesse de l'égouttement au moyen de techniques d'estimation de fréquences. L'appareil mécanique commande la vitesse d'égouttement par changement de l'épaisseur ou du diamètre du tube IV en fonction de la vitesse d'égouttement IV mesurée par le traitement d'image ou de vidéo. Le système d'éclairage éclaire la chambre d'égouttement de manière qu'une image claire puisse être capturée pour le traitement d'image ou de vidéo.
PCT/IB2012/050434 2011-02-02 2012-01-31 Traitement d'image, estimation de fréquence, commande mécanique et éclairage destiné à un système de commande et de surveillance tv automatique Ceased WO2012104779A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201280016765.3A CN104010676A (zh) 2011-02-02 2012-01-31 自动iv监测与控制系统的图像处理、频率估算、机械控制、以及照明

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
US13/019,698 US20120197185A1 (en) 2011-02-02 2011-02-02 Electromechanical system for IV control
US13/019,698 2011-02-02
US13/356,632 US20140327759A1 (en) 2012-01-23 2012-01-23 Image Processing, Frequency Estimation, Mechanical Control and Illumination for an Automatic IV Monitoring and Controlling system
US13/356,632 2012-01-23

Publications (1)

Publication Number Publication Date
WO2012104779A1 true WO2012104779A1 (fr) 2012-08-09

Family

ID=46602126

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/IB2012/050434 Ceased WO2012104779A1 (fr) 2011-02-02 2012-01-31 Traitement d'image, estimation de fréquence, commande mécanique et éclairage destiné à un système de commande et de surveillance tv automatique

Country Status (2)

Country Link
CN (1) CN104010676A (fr)
WO (1) WO2012104779A1 (fr)

Cited By (28)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2014160249A1 (fr) * 2013-03-14 2014-10-02 Baxter International Inc. Chambre compte-gouttes ayant une surface intérieure hydrophobe
WO2014187432A1 (fr) * 2013-05-20 2014-11-27 Tao Kai Moniteur vidéo d'administration intraveineuse
US9199036B2 (en) 2013-02-25 2015-12-01 Shift Labs, Inc. Device, method, and system for monitoring the delivery of fluids through a drip chamber
US9372486B2 (en) 2011-12-21 2016-06-21 Deka Products Limited Partnership System, method, and apparatus for monitoring, regulating, or controlling fluid flow
US9435455B2 (en) 2011-12-21 2016-09-06 Deka Products Limited Partnership System, method, and apparatus for monitoring, regulating, or controlling fluid flow
US9724467B2 (en) 2011-12-21 2017-08-08 Deka Products Limited Partnership Flow meter
US9746093B2 (en) 2011-12-21 2017-08-29 Deka Products Limited Partnership Flow meter and related system and apparatus
US9746094B2 (en) 2011-12-21 2017-08-29 Deka Products Limited Partnership Flow meter having a background pattern with first and second portions
US9759343B2 (en) 2012-12-21 2017-09-12 Deka Products Limited Partnership Flow meter using a dynamic background image
USD799025S1 (en) 2013-11-06 2017-10-03 Deka Products Limited Partnership Apparatus to control fluid flow through a tube
USD802118S1 (en) 2013-11-06 2017-11-07 Deka Products Limited Partnership Apparatus to control fluid flow through a tube
USD813376S1 (en) 2013-11-06 2018-03-20 Deka Products Limited Partnership Apparatus to control fluid flow through a tube
USD815730S1 (en) 2013-11-06 2018-04-17 Deka Products Limited Partnership Apparatus to control fluid flow through a tube
USD816829S1 (en) 2013-11-06 2018-05-01 Deka Products Limited Partnership Apparatus to control fluid flow through a tube
CN105142694B (zh) * 2013-03-14 2018-08-31 巴克斯特国际公司 具有疏水的内表面的滴注器
US10088346B2 (en) 2011-12-21 2018-10-02 Deka Products Limited Partnership System, method, and apparatus for monitoring, regulating, or controlling fluid flow
US10228683B2 (en) 2011-12-21 2019-03-12 Deka Products Limited Partnership System, method, and apparatus for monitoring, regulating, or controlling fluid flow
USD854145S1 (en) 2016-05-25 2019-07-16 Deka Products Limited Partnership Apparatus to control fluid flow through a tube
US10488848B2 (en) 2011-12-21 2019-11-26 Deka Products Limited Partnership System, method, and apparatus for monitoring, regulating, or controlling fluid flow
USD905848S1 (en) 2016-01-28 2020-12-22 Deka Products Limited Partnership Apparatus to control fluid flow through a tube
USD964563S1 (en) 2019-07-26 2022-09-20 Deka Products Limited Partnership Medical flow clamp
US11464905B2 (en) 2013-02-25 2022-10-11 Shift Labs, Inc. Monitoring device including an emitter emitting electromagnetic radiation and a detector positioned to receive the radiation to determine one or more rolling average flow rates
CN115308121A (zh) * 2022-07-28 2022-11-08 重庆川仪自动化股份有限公司 采用图像识别技术测定金属塑性成形摩擦系数的方法
US11583630B2 (en) 2010-10-19 2023-02-21 Baxter International Inc. Optical imaging system with multiple imaging channel optical sensing
CN116205878A (zh) * 2023-02-24 2023-06-02 中国舰船研究设计中心 一种丝网印刷母版灰度图像的几何参数快速提取方法
US11744935B2 (en) 2016-01-28 2023-09-05 Deka Products Limited Partnership Apparatus for monitoring, regulating, or controlling fluid flow
US11839741B2 (en) 2019-07-26 2023-12-12 Deka Products Limited Partneship Apparatus for monitoring, regulating, or controlling fluid flow
CN121095238A (zh) * 2025-11-07 2025-12-09 西安锐博生物科技有限公司 一种丝兰提取物分离过程的实时图像监测方法及系统

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE112016000229T5 (de) * 2015-01-13 2017-09-07 Murata Manufacturing Co., Ltd. Tropfraten-Messvorrichtung, Tropfraten-Controller, Tropfinfusionsvorrichtung, und Flüssigkeitstropfen-Volumenmessvorrichtung
JP7154279B2 (ja) * 2017-08-11 2022-10-17 エフ・ホフマン－ラ・ロシュ・アクチェンゲゼルシャフト 医療デバイスのためのレンチキュラーラベル
CN109568724B (zh) * 2018-11-14 2021-04-02 南京大学 一种基于rfid无源感知的输液滴速监测系统及其工作方法
CN110141724B (zh) * 2019-06-04 2024-08-30 苏州慧尔生医疗科技有限公司 一种输液器滴壶的滴速监测装置
CN112540992B (zh) * 2020-12-16 2024-05-24 辛集市气象局 夏玉米水分亏缺指数数据综合显示系统
CN119314232A (zh) * 2024-12-16 2025-01-14 深圳市讯方技术股份有限公司 自动考勤方法、装置、终端设备和存储介质

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6159186A (en) * 1998-03-13 2000-12-12 Wft Projects (Proprietary) Limited Infusion delivery system
KR20050039780A (ko) * 2005-03-12 2005-04-29 강주완 영상 신호 처리를 이용한 링거액 유량 측정 시스템
CN2808160Y (zh) * 2005-07-06 2006-08-23 重庆邮电学院 输液监控器
CN101201605A (zh) * 2007-12-07 2008-06-18 华中科技大学 一种基于机器视觉的医疗输液速度监测与控制系统
CN201110955Y (zh) * 2007-12-07 2008-09-03 华中科技大学 一种基于机器视觉的医疗输液速度监测与控制装置
CN201379841Y (zh) * 2008-06-26 2010-01-13 吴秀华 网络化视、音频输液监视系统
CN201551688U (zh) * 2009-12-17 2010-08-18 费凡 输液监护系统

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN201666540U (zh) * 2010-04-08 2010-12-08 上海理工大学 多镜面无影灯反射器

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6159186A (en) * 1998-03-13 2000-12-12 Wft Projects (Proprietary) Limited Infusion delivery system
KR20050039780A (ko) * 2005-03-12 2005-04-29 강주완 영상 신호 처리를 이용한 링거액 유량 측정 시스템
CN2808160Y (zh) * 2005-07-06 2006-08-23 重庆邮电学院 输液监控器
CN101201605A (zh) * 2007-12-07 2008-06-18 华中科技大学 一种基于机器视觉的医疗输液速度监测与控制系统
CN201110955Y (zh) * 2007-12-07 2008-09-03 华中科技大学 一种基于机器视觉的医疗输液速度监测与控制装置
CN201379841Y (zh) * 2008-06-26 2010-01-13 吴秀华 网络化视、音频输液监视系统
CN201551688U (zh) * 2009-12-17 2010-08-18 费凡 输液监护系统

Cited By (57)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11583630B2 (en) 2010-10-19 2023-02-21 Baxter International Inc. Optical imaging system with multiple imaging channel optical sensing
US9724465B2 (en) 2011-12-21 2017-08-08 Deka Products Limited Partnership Flow meter
US10894638B2 (en) 2011-12-21 2021-01-19 Deka Products Limited Partnership System, method, and apparatus for monitoring, regulating, or controlling fluid flow
US11793928B2 (en) 2011-12-21 2023-10-24 Deka Products Limited Partnership Flow meter and related method
US9372486B2 (en) 2011-12-21 2016-06-21 Deka Products Limited Partnership System, method, and apparatus for monitoring, regulating, or controlling fluid flow
US9435455B2 (en) 2011-12-21 2016-09-06 Deka Products Limited Partnership System, method, and apparatus for monitoring, regulating, or controlling fluid flow
US9724467B2 (en) 2011-12-21 2017-08-08 Deka Products Limited Partnership Flow meter
US10488848B2 (en) 2011-12-21 2019-11-26 Deka Products Limited Partnership System, method, and apparatus for monitoring, regulating, or controlling fluid flow
US9724466B2 (en) 2011-12-21 2017-08-08 Deka Products Limited Partnership Flow meter
US9746093B2 (en) 2011-12-21 2017-08-29 Deka Products Limited Partnership Flow meter and related system and apparatus
US9746094B2 (en) 2011-12-21 2017-08-29 Deka Products Limited Partnership Flow meter having a background pattern with first and second portions
US10436342B2 (en) 2011-12-21 2019-10-08 Deka Products Limited Partnership Flow meter and related method
US9772044B2 (en) 2011-12-21 2017-09-26 Deka Products Limited Partnership Flow metering using a difference image for liquid parameter estimation
US11738143B2 (en) 2011-12-21 2023-08-29 Deka Products Limited Partnership Flow meier having a valve
US12100507B2 (en) 2011-12-21 2024-09-24 Deka Products Limited Partnership System, method, and apparatus for monitoring, regulating, or controlling fluid flow
US10718445B2 (en) 2011-12-21 2020-07-21 Deka Products Limited Partnership Flow meter having a valve
US9856990B2 (en) 2011-12-21 2018-01-02 Deka Products Limited Partnership Flow metering using a difference image for liquid parameter estimation
US11574407B2 (en) 2011-12-21 2023-02-07 Deka Products Limited Partnership System, method, and apparatus for monitoring, regulating, or controlling fluid flow
US11449037B2 (en) 2011-12-21 2022-09-20 Deka Products Limited Partnership System, method, and apparatus for monitoring, regulating, or controlling fluid flow
US11339887B2 (en) 2011-12-21 2022-05-24 Deka Products Limited Partnership Flow meter and related method
US10739759B2 (en) 2011-12-21 2020-08-11 Deka Products Limited Partnership System, method, and apparatus for monitoring, regulating, or controlling fluid flow
US9976665B2 (en) 2011-12-21 2018-05-22 Deka Products Limited Partnership Flow meter
US10876868B2 (en) 2011-12-21 2020-12-29 Deka Products Limited Partnership System, method, and apparatus for monitoring, regulating, or controlling fluid flow
US10088346B2 (en) 2011-12-21 2018-10-02 Deka Products Limited Partnership System, method, and apparatus for monitoring, regulating, or controlling fluid flow
US10113660B2 (en) 2011-12-21 2018-10-30 Deka Products Limited Partnership Flow meter
US10228683B2 (en) 2011-12-21 2019-03-12 Deka Products Limited Partnership System, method, and apparatus for monitoring, regulating, or controlling fluid flow
US10844970B2 (en) 2011-12-21 2020-11-24 Deka Products Limited Partnership Flow meter
US9759343B2 (en) 2012-12-21 2017-09-12 Deka Products Limited Partnership Flow meter using a dynamic background image
US11464905B2 (en) 2013-02-25 2022-10-11 Shift Labs, Inc. Monitoring device including an emitter emitting electromagnetic radiation and a detector positioned to receive the radiation to determine one or more rolling average flow rates
US9199036B2 (en) 2013-02-25 2015-12-01 Shift Labs, Inc. Device, method, and system for monitoring the delivery of fluids through a drip chamber
US9801996B2 (en) 2013-03-14 2017-10-31 Baxter International Inc. Drip chamber with hydrophobic interior surface
WO2014160249A1 (fr) * 2013-03-14 2014-10-02 Baxter International Inc. Chambre compte-gouttes ayant une surface intérieure hydrophobe
US10314972B2 (en) 2013-03-14 2019-06-11 Baxter International Inc. Drip chamber with hydrophobic interior surface
US9352081B2 (en) 2013-03-14 2016-05-31 Baxter International Inc. Drip chamber with hydrophobic interior surface
CN105142694B (zh) * 2013-03-14 2018-08-31 巴克斯特国际公司 具有疏水的内表面的滴注器
EP3228342A1 (fr) * 2013-03-14 2017-10-11 Baxter International Inc Chambre de goutte à goutte ayant une surface intérieure hydrophobe
US11013860B2 (en) 2013-03-14 2021-05-25 Baxter International Inc. Drip chamber with hydrophobic interior surface
WO2014187432A1 (fr) * 2013-05-20 2014-11-27 Tao Kai Moniteur vidéo d'administration intraveineuse
USD799025S1 (en) 2013-11-06 2017-10-03 Deka Products Limited Partnership Apparatus to control fluid flow through a tube
USD816829S1 (en) 2013-11-06 2018-05-01 Deka Products Limited Partnership Apparatus to control fluid flow through a tube
USD815730S1 (en) 2013-11-06 2018-04-17 Deka Products Limited Partnership Apparatus to control fluid flow through a tube
USD813376S1 (en) 2013-11-06 2018-03-20 Deka Products Limited Partnership Apparatus to control fluid flow through a tube
USD802118S1 (en) 2013-11-06 2017-11-07 Deka Products Limited Partnership Apparatus to control fluid flow through a tube
USD905848S1 (en) 2016-01-28 2020-12-22 Deka Products Limited Partnership Apparatus to control fluid flow through a tube
US11744935B2 (en) 2016-01-28 2023-09-05 Deka Products Limited Partnership Apparatus for monitoring, regulating, or controlling fluid flow
US12502476B2 (en) 2016-01-28 2025-12-23 Deka Products Limited Partnership Apparatus for monitoring, regulating, or controlling fluid flow
USD943736S1 (en) 2016-01-28 2022-02-15 Deka Products Limited Partnership Apparatus to control fluid flow through a tube
USD854145S1 (en) 2016-05-25 2019-07-16 Deka Products Limited Partnership Apparatus to control fluid flow through a tube
USD860437S1 (en) 2016-05-25 2019-09-17 Deka Products Limited Partnership Apparatus to control fluid flow through a tube
USD972125S1 (en) 2016-05-25 2022-12-06 Deka Products Limited Partnership Apparatus to control fluid flow through a tube
USD1060608S1 (en) 2016-05-25 2025-02-04 Deka Products Limited Partnership Device to control fluid flow through a tube
USD972718S1 (en) 2016-05-25 2022-12-13 Deka Products Limited Partnership Apparatus to control fluid flow through a tube
USD964563S1 (en) 2019-07-26 2022-09-20 Deka Products Limited Partnership Medical flow clamp
US11839741B2 (en) 2019-07-26 2023-12-12 Deka Products Limited Partneship Apparatus for monitoring, regulating, or controlling fluid flow
CN115308121A (zh) * 2022-07-28 2022-11-08 重庆川仪自动化股份有限公司 采用图像识别技术测定金属塑性成形摩擦系数的方法
CN116205878A (zh) * 2023-02-24 2023-06-02 中国舰船研究设计中心 一种丝网印刷母版灰度图像的几何参数快速提取方法
CN121095238A (zh) * 2025-11-07 2025-12-09 西安锐博生物科技有限公司 一种丝兰提取物分离过程的实时图像监测方法及系统

Also Published As

Publication number Publication date
CN104010676A (zh) 2014-08-27

Similar Documents

Publication Publication Date Title
WO2012104779A1 (fr) Traitement d'image, estimation de fréquence, commande mécanique et éclairage destiné à un système de commande et de surveillance tv automatique
US20140327759A1 (en) Image Processing, Frequency Estimation, Mechanical Control and Illumination for an Automatic IV Monitoring and Controlling system
Xu et al. Reconstructions in limited‐view thermoacoustic tomography
Matsuda et al. Finger-vein authentication based on deformation-tolerant feature-point matching
Chikkerur et al. Fingerprint image enhancement using STFT analysis
US20100079385A1 (en) Method for calibrating an interactive input system and interactive input system executing the calibration method
Li et al. An adaptive method of speckle reduction and feature enhancement for SAR images based on curvelet transform and particle swarm optimization
Lee et al. Skewed rotation symmetry group detection
Chen et al. 2D finite rate of innovation reconstruction method for step edge and polygon signals in the presence of noise
Mujica et al. A new motion parameter estimation algorithm based on the continuous wavelet transform
Budianto et al. Marker encoded fringe projection profilometry for efficient 3D model acquisition
Zhao et al. The fourier-argand representation: An optimal basis of steerable patterns
US20180365810A1 (en) Object image recovery from digital holograms
Rangayyan et al. Polygonal approximation of contours based on the turning angle function
Noh et al. Touchless sensor capturing five fingerprint images by one rotating camera
Kang et al. A study on fast iris restoration based on focus checking
Elouedi et al. The polynomial discrete Radon transform
Moniruzzaman et al. Face recognition using 4-PSK joint transform correlation
US12608973B2 (en) Scanning device using high-resolution moisture discriminating optics
RU2314027C2 (ru) Устройство для регистрации рисунка кожных линий
Kirkegaard et al. An S factor analysis on the provinces of Vietnam: Relationships with cognitive ability, ethnicity, and Latitude
Sastry et al. A convolution back projection algorithm for local tomography
Gnanaprakasam et al. Average-half-face in 2D and 3D using wavelets for face recognition
Shen et al. Multiscale Feature Attribution for Outliers
김범수 et al. Partial Denoising Boundary Image Matching Based on Time-Series Data

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: 12741497

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 12741497

Country of ref document: EP

Kind code of ref document: A1