Dual-foot sole video evaluation method, device, terminal device and storage medium based on plantar perfusion imaging

Abstract

This application relates to the field of biomedical engineering technology. This application discloses a method, device, terminal device and storage medium for dual-foot sole video evaluation based on plantar perfusion imaging, which can improve the accuracy of evaluation results, simplify the measurement process and enhance the patient experience. The method includes obtaining a dual-foot sole video; obtaining a plurality of left plantar rPPG signals and a plurality of right plantar rPPG signals based on the dual-foot sole video; drawing a plurality of dual-foot sole perfusion maps based on the plurality of left plantar rPPG signals and the plurality of right plantar rPPG signals; generating a target image based on the plurality of dual-foot sole perfusion maps; extracting the average value, standard deviation, maximum value and minimum value of the blood perfusion index ratio from the target image and inputting them into a classifier for evaluation processing to obtain an evaluation result, and medical staff refer to the evaluation result to diagnose peripheral artery disease and its severity.

Description

Technical Field

[0001] This application relates to the field of biomedical engineering technology. More specifically, this application relates to a method, device, terminal device, and storage medium for dual-foot sole video evaluation based on plantar perfusion imaging. Background Art

[0002] For traditional peripheral artery disease evaluation methods, refer to the patent application document with the publication number CN112587119A. This document collects tissue blood flow change data of the gastrocnemius muscle of peripheral artery disease patients and healthy volunteers through a diffusion correlation spectroscopy blood flow detection system; during measurement, the tested subjects are all in a lying position, the probe is placed on the gastrocnemius muscle of each tested subject, and the cuff is placed at the root of the thigh of the lower limb to be measured by the tested subject to block blood flow; for this tissue blood flow change data, after normalization preprocessing, the tissue blood flow change data of the tested subjects is randomly divided into two groups: a training set and a test set; then, the training set samples are data-augmented through a data augmentation method, and the training set samples after data augmentation will be used as the data input of the deep learning network; the data of the augmented training set samples is input into a dual-view convolutional recurrent neural network (dual-view CRNN, dvCRNN) for training. After training is completed, the test set sample data is input into the trained network model for recognition to obtain an evaluation result (i.e., the diagnostic result in this document). When using this method for measurement, a cuff needs to be placed at the root of the thigh of the lower limb to be measured by the tested subject to block blood flow. This operation not only complicates the measurement process but also causes discomfort to the tested subject due to the pressure exerted by the cuff at the root of the thigh of the tested subject, reducing the experience of the tested subject (i.e., the patient).

[0003] In addition, the part where this method collects data is the gastrocnemius muscle of the human body. The internal blood flow state of this part may vary due to differences in muscle depth and position. If the diffusion correlation spectroscopy blood flow detection system fails to capture these subtle changes during the process of collecting tissue blood flow change data, the data collected by the diffusion correlation spectroscopy blood flow detection system may not comprehensively reflect the blood flow state of the gastrocnemius muscle, resulting in the trained network model being unable to accurately obtain the evaluation result of peripheral artery disease. Summary of the Invention

[0004] The purpose of the embodiments of this application is to provide a method, device, terminal device, and storage medium for dual-foot sole video evaluation based on plantar perfusion imaging, which can improve the accuracy of the evaluation result, simplify the measurement process, and improve the patient experience. The embodiments of this application are mainly implemented through the following technical solutions:

[0005] In the first aspect of the embodiments of this application, a method for dual-foot sole video evaluation based on plantar perfusion imaging is provided, including:

[0006] Obtain the video of the soles of the patient's feet;

[0007] Preprocess the video of the soles of the feet to obtain at least one first pixel time-domain change signal and at least one second pixel time-domain change signal;

[0008] Perform division and denoising processing on each first pixel time-domain change signal to obtain a plurality of left sole rPPG signals corresponding to each first pixel time-domain change signal;

[0009] Perform division and denoising processing on each second pixel time-domain change signal to obtain a plurality of right sole rPPG signals corresponding to each second pixel time-domain change signal;

[0010] Draw a plurality of perfusion maps of the soles of the feet based on the plurality of left sole rPPG signals and the plurality of right sole rPPG signals;

[0011] Generate a target image based on the plurality of perfusion maps of the soles of the feet;

[0012] Extract the average value, standard deviation, maximum value, and minimum value of the blood perfusion index ratio from the target image and input them into a classifier for evaluation processing to obtain the evaluation result of the video of the soles of the patient's feet.

[0013] According to an embodiment of the present application, the step of preprocessing the video of the soles of the feet to obtain at least one first pixel time-domain change signal and at least one second pixel time-domain change signal includes:

[0014] Perform decomposition processing on the video of the soles of the feet to obtain multiple frames of sole images;

[0015] Perform average pooling processing on each frame of sole image to obtain a to-be-processed image corresponding to each frame of sole image;

[0016] Identify the sole part on each to-be-processed image as the region of interest corresponding to each to-be-processed image;

[0017] Obtain the first RGB three-channel pixel values of each pixel point in the left foot region in each region of interest, and obtain the second RGB three-channel pixel values of each pixel point in the right foot region in each region of interest;

[0018] Based on all the first RGB three-channel pixel values, obtain the first pixel time-domain change signal corresponding to each pixel point in the left foot region according to the time change;

[0019] Based on all the second RGB three-channel pixel values, obtain the second pixel time-domain change signal corresponding to each pixel point in the right foot region according to the time change.

[0020] According to an embodiment of the present application, the steps of dividing and denoising each first pixel time-domain variation signal to obtain a plurality of left-sole rPPG signals corresponding to each first pixel time-domain variation signal include:

[0021] Set the sliding window length and step size;

[0022] Based on the sliding window length and the step size, perform division processing on each first pixel time-domain variation signal to obtain a plurality of first sub-signals corresponding to each first pixel time-domain variation signal;

[0023] Use a Butterworth filter to perform filtering processing on each first sub-signal to obtain a left-sole rPPG signal corresponding to each first sub-signal.

[0024] According to an embodiment of the present application, the steps of drawing a plurality of dual-sole perfusion maps based on the plurality of left-sole rPPG signals and the plurality of right-sole rPPG signals include:

[0025] Use a peak detection algorithm to detect the first peaks and first valleys of each channel in the target left-sole rPPG signal, where the target left-sole rPPG signal is any one of the plurality of left-sole rPPG signals;

[0026] Use the average value of the amplitude differences between the first peaks and the first valleys of each channel as the first alternating current volume of the target left-sole rPPG signal;

[0027] Use the average value of the overall amplitude of the target left-sole rPPG signal as the first direct current volume of the target left-sole rPPG signal;

[0028] Calculate the first blood perfusion index corresponding to the target left-sole rPPG signal using the first alternating current volume and the first direct current volume;

[0029] Use the peak detection algorithm to detect the second peaks and second valleys of each channel in the target right-sole rPPG signal, where the target right-sole rPPG signal is any one of the plurality of right-sole rPPG signals;

[0030] Use the average value of the amplitude differences between the second peaks and the second valleys of each channel as the second alternating current volume of the target right-sole rPPG signal;

[0031] Use the average value of the overall amplitude of the target right-sole rPPG signal as the second direct current volume of the target right-sole rPPG signal;

[0032] Calculate the second blood perfusion index corresponding to the target right-sole rPPG signal using the second alternating current volume and the second direct current volume;

[0033] Plot the multiple bipedal plantar perfusion maps based on all the first blood perfusion indices and all the second blood perfusion indices.

[0034] According to an embodiment of the present application, the step of generating a target image based on the multiple bipedal plantar perfusion maps includes:

[0035] Calculate a first spatial average value of the first blood perfusion index in each bipedal plantar perfusion map;

[0036] Calculate a second spatial average value of the second blood perfusion index in each bipedal plantar perfusion map;

[0037] Connect all the first spatial average values to form a left-foot blood perfusion index change wave;

[0038] Connect all the second spatial average values to form a right-foot blood perfusion index change wave;

[0039] Generate the target image based on the ratio of the left-foot blood perfusion index change wave and the right-foot blood perfusion index change wave.

[0040] According to an embodiment of the present application, the bipedal plantar video evaluation method based on plantar perfusion imaging further includes a training step of the classifier, and the training step of the classifier includes:

[0041] Obtain a training data set and a true label set, and each training data in the training data set has a one-to-one correspondence with one of the true labels in the true label set;

[0042] Perform feature extraction processing on target training data to obtain training features, where the target training data is any training data in the training data set;

[0043] Input the training features into an original classifier for evaluation processing to obtain a prediction result;

[0044] Calculate a loss function based on the prediction result and the true label corresponding to the target training data;

[0045] Adjust the parameters of the original classifier based on the loss function to obtain the classifier.

[0046] In a second aspect of the embodiments of the present application, there is provided a bipedal plantar video evaluation device based on plantar perfusion imaging, including:

[0047] A bipedal plantar video acquisition module, configured to acquire a bipedal plantar video of a patient;

[0048] A preprocessing module for preprocessing the dual-foot sole video to obtain at least one first pixel time-domain change signal and at least one second pixel time-domain change signal;

[0049] A left-foot sole rPPG signal acquisition module for dividing and denoising each first pixel time-domain change signal to obtain a plurality of left-foot sole rPPG signals corresponding to each first pixel time-domain change signal;

[0050] A right-foot sole rPPG signal acquisition module for dividing and denoising each second pixel time-domain change signal to obtain a plurality of right-foot sole rPPG signals corresponding to each second pixel time-domain change signal;

[0051] A drawing module for drawing a plurality of dual-foot sole perfusion maps based on the plurality of left-foot sole rPPG signals and the plurality of right-foot sole rPPG signals;

[0052] A generation module for generating a target image based on the plurality of dual-foot sole perfusion maps;

[0053] An evaluation module for extracting the average value, standard deviation, maximum value, and minimum value of the blood perfusion index ratio from the target image and inputting them into a classifier for evaluation processing to obtain an evaluation result of the dual-foot sole video of the patient.

[0054] In the third aspect of the embodiments of the present application, a dual-foot sole video evaluation device based on plantar perfusion imaging is provided, including: a terminal device, an optical imaging device, a fixing bracket, and a transparent plate. The optical imaging device includes a camera and an LED fill light;

[0055] The terminal device includes a processor, and the processor is used to execute the steps of the dual-foot sole video evaluation method provided in the first aspect of the embodiments of the present application;

[0056] The transparent plate is placed on the upper part of the fixing bracket. During the process of the camera recording the dual-foot sole video of the patient, the dual feet of the patient are placed on the transparent plate, and the LED fill light irradiates the transparent plate;

[0057] The optical imaging device is placed inside the fixing bracket;

[0058] After the camera finishes recording the dual-foot sole video, it sends the dual-foot sole video to the processor.

[0059] In a fourth aspect of the embodiments of the present application, a terminal device is provided, including: a processor and a memory. The memory is used to store a computer program, and the processor is used to call and run the computer program stored in the memory to execute the steps of the method for evaluating a dual-foot sole video based on plantar perfusion imaging provided in the first aspect of the embodiments of the present application.

[0060] In a fifth aspect of the embodiments of the present application, a computer-readable storage medium is provided. The computer-readable storage medium is used to store a computer program, and the computer program causes a computer to execute the steps of the method for evaluating a dual-foot sole video based on plantar perfusion imaging provided in the first aspect of the embodiments of the present application.

[0061] The beneficial effects of the embodiments of the present application include:

[0062] In the embodiments of the present application, a pulse wave signal is extracted from a dual-foot sole video. Then, a blood perfusion index of both feet is calculated based on the pulse wave signal. Furthermore, a dual-foot sole perfusion map is generated through the blood perfusion index to analyze the distribution difference of the blood perfusion index of both feet in space and time, so as to obtain an evaluation result of the dual-foot sole video, enabling medical staff to diagnose peripheral artery disease and the severity of peripheral artery disease according to the evaluation result. Specifically, in the embodiments of the present application, a dual-foot sole video of a patient is acquired; the dual-foot sole video is preprocessed to obtain at least one first pixel time-domain change signal and at least one second pixel time-domain change signal; each first pixel time-domain change signal is divided and denoised to obtain a plurality of left-foot sole rPPG signals corresponding to each first pixel time-domain change signal; each second pixel time-domain change signal is divided and denoised to obtain a plurality of right-foot sole rPPG signals corresponding to each second pixel time-domain change signal; a plurality of dual-foot sole perfusion maps are drawn based on the plurality of left-foot sole rPPG signals and the plurality of right-foot sole rPPG signals; a target image is generated based on the plurality of dual-foot sole perfusion maps; an average value, a standard deviation, a maximum value, and a minimum value of a blood perfusion index ratio are extracted from the target image and input into a classifier for evaluation processing to obtain an evaluation result of the dual-foot sole video of the patient. Thus, compared with the prior art, the embodiments of the present application do not require an additional instrument such as a cuff to assist in measuring data, thereby simplifying the measurement process and improving the patient experience. Moreover, in the embodiments of the present application, the dual-foot soles of the human body are used as the data acquisition part, which can reduce the influence of changes in blood flow state due to different muscle depths and positions, thereby improving the accuracy of the evaluation result, enabling medical staff to accurately evaluate the condition of lower limb atherosclerosis. Description of the Drawings

[0063] To more clearly illustrate the technical solutions in the embodiments of the present application or in the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0064] Figure 1 It is a flowchart of the method for evaluating the video of the soles of both feet based on plantar perfusion imaging in some embodiments of the present application;

[0065] Figure 2 It is a flowchart of the method for evaluating the video of the soles of both feet based on plantar perfusion imaging in other embodiments of the present application;

[0066] Figure 3 It is a reference diagram of the region of interest corresponding to the sole of the foot without pressure;

[0067] Figure 4 It is a reference diagram of the region of interest corresponding to the sole of the foot with pressure;

[0068] Figure 5 It is a reference diagram of the target left sole rPPG signal corresponding to the sole of the foot without pressure;

[0069] Figure 6 It is a reference diagram of the target left sole rPPG signal corresponding to the sole of the foot with pressure;

[0070] Figure 7 It is a perfusion map of the soles of both feet corresponding to the sole of the foot without pressure;

[0071] Figure 8 It is a perfusion map of the soles of both feet corresponding to the sole of the foot with pressure;

[0072] Figure 9 It is an ultrasonic waveform diagram corresponding to 0% systolic blood pressure;

[0073] Figure 10 It is an ultrasonic waveform diagram corresponding to 40% systolic blood pressure;

[0074] Figure 11 It is an ultrasonic waveform diagram corresponding to 80% systolic blood pressure;

[0075] Figure 12 It is an ultrasonic waveform diagram corresponding to 120% systolic blood pressure;

[0076] Figure 13 For Figure 9 The perfusion map of the soles of both feet corresponding to the ultrasonic waveform in

[0077] Figure 14 For Figure 10Dual-foot sole perfusion images corresponding to the middle ultrasonic waveform;

[0078] Figure 15 is Figure 11 Dual-foot sole perfusion images corresponding to the middle ultrasonic waveform;

[0079] Figure 16 is Figure 12 Dual-foot sole perfusion images corresponding to the middle ultrasonic waveform;

[0080] Figure 17 is the principle block diagram of the dual-foot sole video evaluation device based on sole perfusion imaging in some embodiments of the present application;

[0081] Figure 18 is the principle block diagram of the dual-foot sole video evaluation device based on sole perfusion imaging in some other embodiments of the present application;

[0082] Figure 19 is the application scenario diagram of the sole imager and medical ultrasonic instrument of the present application in some embodiments;

[0083] Figure 20 is the principle block diagram of the terminal device of the present application in some embodiments. Detailed implementation manners

[0084] To make the above objects, features, and advantages of the present application more obvious and understandable, the following describes the detailed implementation manners of the present application with reference to the accompanying drawings. Many specific details are set forth in the following description to fully understand the present application. However, the present application can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without departing from the connotation of the present application. Therefore, the present application is not limited by the specific embodiments disclosed below.

[0085] It should be noted that the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include at least one of such features. In the description of the present application, "a plurality" means at least two, such as two, three, etc., unless otherwise specifically defined.

[0086] The term "exemplary" or "for example" and the like are used to indicate examples, illustrations, or explanations. Any embodiment or design solution described as "exemplary" or "for example" in the embodiments of the present application should not be construed as being more preferred or having more advantages than other embodiments or design solutions. Rather, the use of the term "exemplary" or "for example" and the like is intended to present the relevant concepts in a specific manner.

[0087] The term "comprising", "including" or any other variation thereof is intended to cover non-exclusive inclusion, e.g., a process, method, system, product or device that comprises a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these process, method, product or device.

[0088] Unless otherwise defined, all technical and scientific terms used in the specification of this application have the same meaning as commonly understood by those skilled in the art to which this application pertains. The terms used in the specification of this application are for the purpose of describing specific embodiments only and are not intended to limit this application. The term "and / or" used in the specification of this application includes any and all combinations of one or more of the related listed items.

[0089] The term "Peripheral Artery Disease (PAD)" is a common cardiovascular disease, which is mainly manifested as peripheral atherosclerosis. This is a chronic, progressive vascular lesion that mainly occurs in the intimal layer of medium and large arteries. Atherosclerosis is characterized by lipid deposition and fibrous tissue hyperplasia on the arterial intima, resulting in vascular stenosis or occlusion.

[0090] The occurrence of PAD is usually related to multiple factors, such as smoking, diabetes, obesity, hypertension, and irregular work and rest. Timely and accurate assessment of the arterial vascular health status and detection of the presence of atherosclerosis can improve the quality of life of patients, relieve the anxiety of patients, and avoid the high treatment costs in the late stage of PAD.

[0091] Traditional methods for evaluating peripheral artery disease involve collecting tissue blood flow change data from the gastrocnemius of patients with peripheral artery disease and healthy volunteers using a diffusion correlation spectroscopy blood flow detection system. During the measurement, the subjects being tested are all in a lying position. The probe is placed on the gastrocnemius of each subject, and a cuff is placed at the root of the thigh of the lower limb being measured to block blood flow. For this tissue blood flow change data, after normalization preprocessing, the tissue blood flow change data of the subjects being tested are randomly divided into two groups: a training set and a test set. Then, the training set samples are augmented through data augmentation methods, and the augmented training set samples will be used as the data input for the deep learning network. The augmented training set sample data are input into a dual-view convolutional recurrent neural network (dual-view CRNN, dvCRNN) for training. After training is completed, the test set sample data are input into the trained network model for recognition to obtain the evaluation result (i.e., the diagnostic result in this literature). When using this method for measurement, a cuff needs to be placed at the root of the thigh of the lower limb being measured to block blood flow. This operation not only complicates the measurement process but also causes discomfort to the subject being tested due to the pressure exerted by the cuff at the root of the thigh, reducing the experience of the subject being tested (i.e., the patient). In addition, the location where this method collects data is the gastrocnemius of the human body. The internal blood flow state of this location may vary due to differences in muscle depth and position. If the diffusion correlation spectroscopy blood flow detection system fails to capture these subtle changes during the process of collecting tissue blood flow change data, then the data collected by the diffusion correlation spectroscopy blood flow detection system may not comprehensively reflect the blood flow state of the gastrocnemius, resulting in the trained network model being unable to accurately obtain the evaluation result of peripheral artery disease. Based on this, the present application provides a dual-foot sole video evaluation method based on plantar perfusion imaging to solve the above problems.

[0092] The following further describes the specific embodiments of the present application with reference to the accompanying drawings.

[0093] Reference Figure 1 As shown, it is a flowchart of a dual-foot sole video evaluation method based on plantar perfusion imaging provided in the first aspect of the embodiment of the present application. In Figure 1 it, the dual-foot sole video evaluation method based on plantar perfusion imaging includes:

[0094] S1. Obtain the dual-foot sole videos of the patient.

[0095] The dual-foot sole videos are obtained by recording with a camera of a dual-foot sole video evaluation device based on plantar perfusion imaging.

[0096] The dual-foot sole videos are RGB videos. The RGB videos refer to video data encoded and represented using the RGB (red, green, blue) color space.

[0097] S2. Preprocess the video of the soles of both feet to obtain at least one first pixel time-domain change signal and at least one second pixel time-domain change signal.

[0098] Further, step S2 includes:

[0099] S21. Decompose the video of the soles of both feet to obtain multiple frames of images of the soles of both feet.

[0100] The decomposition process can be understood as a decoding process.

[0101] In the embodiment of the present application, the image size of each frame of the image of the soles of both feet is 480px×300px (px is the pixel unit). In other embodiments, the image size of each frame of the image of the soles of both feet can be set by those skilled in the art according to actual needs, and no further limitation is made herein.

[0102] S22. Perform average pooling processing on each frame of the image of the soles of both feet to obtain a to-be-processed image corresponding to each frame of the image of the soles of both feet.

[0103] In the embodiment of the present application, a pooling window of 50px×50px is used for average pooling processing. In other embodiments, the specific size of the pooling window can be set by those skilled in the art according to actual needs.

[0104] In other embodiments, methods such as max pooling cooperation and min pooling can be used to replace the average pooling method, or the superpixel image segmentation method can be used to downsample each frame of the image of the soles of both feet to achieve regional pixel fusion.

[0105] The setting of step S22 can downsample the entire video of the soles of both feet to reduce the amount of calculation and weaken the influence of noise on individual pixel points, thereby enhancing the robustness of the method for evaluating the video of the soles of both feet based on plantar perfusion imaging.

[0106] In other embodiments, the method for downsampling the entire video of the soles of both feet can also be the pixel fusion method, the linear interpolation method, or the superpixel method.

[0107] S23. Identify the soles of both feet on each to-be-processed image as the region of interest corresponding to each to-be-processed image. Step S23 can refer to Figure 2 the step of "selecting the region of interest" in Figure 3 and Figure 4 as shown. Among them, Figure 3 is the region of interest corresponding to the sole of the foot without pressure, Figure 4 is the region of interest corresponding to the sole of the foot with pressure.

[0108] Specifically, an object detection algorithm can be used to implement step S23. The object detection algorithm can be R-CNN (Region-based Convolutional Network), Fast R-CNN (Fast Region-based Convolutional Network), or Faster R-CNN (Faster Region-based Convolutional Neural Networks).

[0109] S24. Obtain the first RGB three-channel pixel values of each pixel point in the left foot region in each region of interest, and obtain the second RGB three-channel pixel values of each pixel point in the right foot region in each region of interest.

[0110] It should be understood that the RGB described herein refers to the three colors of red, green, and blue.

[0111] S25. Based on all the first RGB three-channel pixel values, obtain the first pixel time-domain change signal corresponding to each pixel point in the left foot region according to the time change.

[0112] S26. Based on all the second RGB three-channel pixel values, obtain the second pixel time-domain change signal corresponding to each pixel point in the right foot region according to the time change.

[0113] S3. Perform division and denoising processing on each first pixel time-domain change signal to obtain multiple left sole rPPG signals corresponding to each first pixel time-domain change signal. Step S3 can refer to the Figure 2 step of "extracting the left sole rPPG signal" in

[0114] It should be understood that the rPPG (Remote Photo Plethysmo Graphy) described herein refers to remote photoplethysmography.

[0115] Further, step S3 includes:

[0116] S31. Set the sliding window length and the step size.

[0117] The sliding window length is 5 seconds, and the step size is 1 second. In other embodiments, the specific values of the sliding window length and the step size can be set by those skilled in the art according to actual needs.

[0118] S32. Divide each first pixel time-domain change signal based on the sliding window length and the step size to obtain a plurality of first sub-signals corresponding to each first pixel time-domain change signal.

[0119] Each of the first sub-signals is an RGB signal of a sub-region of the corresponding first pixel time-domain change signal.

[0120] S33. Filter each first sub-signal using a Butterworth filter to obtain a left sole rPPG signal corresponding to each first sub-signal.

[0121] The filtering frequency of the Butterworth filter is set to [0.6, 4] Hz to remove noise interference in each first sub-signal.

[0122] The left sole rPPG signal is a three-channel pulse wave signal.

[0123] In the embodiments of the present application, the G (Green) channel signal is selected as the signal to be used.

[0124] S4. Perform division and denoising processing on each second pixel time-domain change signal to obtain a plurality of right sole rPPG signals corresponding to each second pixel time-domain change signal. The steps of S4 can refer to Figure 2 the step of "extracting the right sole rPPG signal" in

[0125] Further, the steps of S4 include:

[0126] S41. Divide each second pixel time-domain change signal based on the sliding window length and the step size to obtain a plurality of second sub-signals corresponding to each second pixel time-domain change signal.

[0127] Each of the second sub-signals is an RGB signal of a sub-region of the corresponding second pixel time-domain change signal.

[0128] S42. Filter each second sub-signal using the Butterworth filter to obtain a right sole rPPG signal corresponding to each second sub-signal.

[0129] The filtering frequency of the Butterworth filter is set to [0.6, 4] Hz to remove noise interference in each second sub-signal.

[0130] The right sole rPPG signal is a three-channel pulse wave signal.

[0131] S5. Draw a plurality of dual-sole perfusion maps based on the plurality of left sole rPPG signals and the plurality of right sole rPPG signals. The steps of S5 can refer to Figure 2 the step of "drawing a plurality of dual-sole perfusion maps" in

[0132] Further, step S5 includes:

[0133] S51. Detect the first peak and the first trough of each channel in the target left - foot rPPG signal by using a peak - detection algorithm, where the target left - foot rPPG signal is any one of the multiple left - foot rPPG signals.

[0134] The peak - detection algorithm is a commonly used technique in signal processing for identifying local maxima or minima in a data sequence.

[0135] Exemplarily, the target left - foot rPPG signal can refer to Figure 5 and Figure 6 As shown, in Figure 5 and Figure 6 the first peak is expressed as "peak", the first trough is expressed as "trough", the first alternating component of the target left - foot rPPG signal is expressed as "AC", and the first direct - current component of the target left - foot rPPG signal is expressed as "DC". Figure 5 is a reference diagram of the target left - foot rPPG signal corresponding to the non - pressured sole. Referring to the target left - foot rPPG signal in Figure 5 the ratio of its first alternating component to its first direct - current component can be 0.52. Figure 6 is a reference diagram of the target left - foot rPPG signal corresponding to the pressured sole. Referring to the target left - foot rPPG signal in Figure 6 the ratio of its first alternating component to its first direct - current component can be 1.98.

[0136] S52. Use the average value of the amplitude difference between the first peak and the first trough of each channel as the first alternating component of the target left - foot rPPG signal.

[0137] S53. Use the average value of the overall amplitude of the target left - foot rPPG signal as the first direct - current component of the target left - foot rPPG signal.

[0138] S54. Calculate the first blood perfusion index corresponding to the target left - foot rPPG signal by using the first alternating component and the first direct - current component.

[0139] Specifically, the calculation formula of S54 is:

[0140] ;

[0141] where is the first blood perfusion index corresponding to the target left - foot rPPG signal; is the first alternating component of the target left - foot rPPG signal; is the first DC component of the target left plantar rPPG signal.

[0142] In other embodiments, the standard deviation of the amplitude of the target left plantar rPPG signal can also be used as the first AC component, and the average value of the amplitude of the target left plantar rPPG signal can be used as the first DC component.

[0143] S55. Detect the second peaks and second valleys of each channel in the target right plantar rPPG signal by using the peak detection algorithm, where the target right plantar rPPG signal is any one of the multiple right plantar rPPG signals.

[0144] S56. Use the average value of the amplitude difference between the second peak and the second valley of each channel as the second AC component of the target right plantar rPPG signal.

[0145] S57. Use the average value of the overall amplitude of the target right plantar rPPG signal as the second DC component of the target right plantar rPPG signal.

[0146] S58. Calculate the second blood perfusion index corresponding to the target right plantar rPPG signal by using the second AC component and the second DC component.

[0147] Specifically, the calculation formula of S58 is:

[0148] ;

[0149] where is the second blood perfusion index corresponding to the target right plantar rPPG signal; is the second AC component of the target right plantar rPPG signal; is the second DC component of the target right plantar rPPG signal.

[0150] In other embodiments, the standard deviation of the amplitude of the target right plantar rPPG signal can also be used as the second AC component, and the average value of the amplitude of the target right plantar rPPG signal can be used as the second DC component.

[0151] Both the first blood perfusion index and the second blood perfusion index are used to reflect the transmission of pulse waves and blood flow conditions in human blood vessels.

[0152] S59. Draw the multiple dual-plantar perfusion maps based on all the first blood perfusion indices and all the second blood perfusion indices. The dual-plantar perfusion maps can refer to Figure 7 and Figure 8 as shown, where Figure 7The perfusion maps of both plantar soles corresponding to the plantar soles without pressure. Figure 8 The perfusion maps of both plantar soles corresponding to the plantar soles with pressure.

[0153] Further, step S59 includes:

[0154] S591: Taking the time period corresponding to the target left plantar rPPG signal and the target right plantar rPPG signal as the target time period.

[0155] S592: Drawing the perfusion maps of both plantar soles corresponding to the target time period by using all the first blood perfusion indices and all the second blood perfusion indices corresponding to the target time period.

[0156] In the process of drawing the perfusion maps of both plantar soles in the embodiments of the present application, a cold color tone is used to represent low perfusion, and a warm color tone is used to represent high perfusion.

[0157] S593: All the perfusion maps of both plantar soles constitute the multiple perfusion maps of both plantar soles.

[0158] S6: Generating a target image based on the multiple perfusion maps of both plantar soles.

[0159] The target image is a change map of the perfusion index (PI) ratio.

[0160] Further, step S6 includes:

[0161] S61: Calculating the first spatial average value of the first blood perfusion index in each perfusion map of both plantar soles.

[0162] S62: Calculating the second spatial average value of the second blood perfusion index in each perfusion map of both plantar soles.

[0163] S63: Connecting all the first spatial average values to form a change wave of the left foot blood perfusion index.

[0164] S64: Connecting all the second spatial average values to form a change wave of the right foot blood perfusion index.

[0165] S65: Generating the target image based on the ratio of the change wave of the left foot blood perfusion index and the change wave of the right foot blood perfusion index. Step S65 can refer to the step of Figure 2 "Calculating the ratio of the change wave of the left foot blood perfusion index and the change wave of the right foot blood perfusion index" in

[0166] The reason for adopting step S65 is that there are significant individual differences between the first blood perfusion index and the second blood perfusion index. In the embodiments of the present application, the method of using the ratio of the left-foot blood perfusion index change wave and the right-foot blood perfusion index change wave is adopted to weaken the individual differences and achieve the identification of PAD (Peripheral Artery Disease) for different patients, thereby obtaining a new target image.

[0167] S7. Extract the average value, standard deviation, maximum value, and minimum value of the blood perfusion index ratio from the target image and input them into a classifier for evaluation processing to obtain the evaluation result of the dual-foot sole video of the patient.

[0168] Medical staff can diagnose peripheral artery disease and the severity of peripheral artery disease based on the evaluation result.

[0169] Step S7 can refer to Figure 2 the "evaluate PAD" step in

[0170] The "blood perfusion index ratio" mentioned in step S7 refers to the ratio of the left-foot blood perfusion index change wave and the right-foot blood perfusion index change wave.

[0171] The classifier is a classifier for supervised learning. Exemplarily, the classifier can be a support vector machine, a random forest, or a multilayer perceptron (MLP).

[0172] In other embodiments, the classifier can be replaced with other machine learning models.

[0173] In the embodiments of the present application, by extracting the pulse wave signal from the dual-foot sole video, then calculating the blood perfusion index of both feet based on the pulse wave signal, and generating a dual-foot sole perfusion map through the blood perfusion index to analyze the spatio-temporal distribution differences of the blood perfusion index of both feet, the evaluation result of the dual-foot sole video is obtained, so that medical staff can diagnose peripheral artery disease and the severity of peripheral artery disease based on the evaluation result. Thus, compared with the prior art, the embodiments of the present application do not need to use an additional device such as a cuff to assist in measuring data, thereby simplifying the measurement process and improving the patient experience. Moreover, the embodiments of the present application use the dual-foot soles of the human body as the data collection site, which can reduce the influence of changes in blood flow state due to different muscle depths and positions, thereby improving the accuracy of the evaluation result, enabling medical staff to accurately evaluate the condition of lower limb atherosclerosis.

[0174] The embodiments of the present application have the characteristics of low cost and high precision, achieving the goal of bringing PAD detection from the hospital to daily life. Moreover, it can evaluate PAD timely and accurately to formulate or adjust the treatment strategy of PAD and delay the development of PAD disease.

[0175] In some embodiments, the method for evaluating the video of both feet based on plantar perfusion imaging further includes the training step of the classifier, and the training step of the classifier includes:

[0176] S71. Obtain a training data set and a true label set, and there is a one-to-one correspondence between each training data in the training data set and one of the true labels in the true label set.

[0177] Each training data in the training data set is a training video of both feet recorded by the camera.

[0178] Each true label in the true label set is a true ultrasonic waveform diagram. Specifically, the ultrasonic waveform diagram can refer to Figure 9 、 Figure 10 、 Figure 11 and Figure 12 as shown. The Figure 9 is the ultrasonic waveform diagram corresponding to 0% systolic blood pressure, Figure 10 is the ultrasonic waveform diagram corresponding to 40% systolic blood pressure, Figure 11 is the ultrasonic waveform diagram corresponding to 80% systolic blood pressure, Figure 12 is the ultrasonic waveform diagram corresponding to 120% systolic blood pressure. Figure 9 The plantar perfusion map of both feet corresponding to the ultrasonic waveform in Figure 13 is Figure 10 The plantar perfusion map of both feet corresponding to the ultrasonic waveform in Figure 14 is Figure 11 The plantar perfusion map of both feet corresponding to the ultrasonic waveform in Figure 15 is Figure 12 The plantar perfusion map of both feet corresponding to the ultrasonic waveform in Figure 16 is Figure 9 、 Figure 10 、 Figure 11 、 Figure 12 、 Figure 13 、 Figure 14 、 Figure 15 and Figure 16 In, L represents the left foot and R represents the right foot.

[0179] The ultrasonic waveform diagram is obtained by a medical ultrasonic instrument, and the medical ultrasonic instrument can refer to the label B in Figure 19 。

[0180] S72. Extract feature from the target training data to obtain training features, where the target training data is any one of the training data in the training dataset.

[0181] Specifically, step S72 includes:

[0182] S721. Preprocess the target training data to obtain at least one first pixel time-domain change training signal and at least one second pixel time-domain change training signal.

[0183] S722. Divide and denoise each first pixel time-domain change training signal to obtain multiple left sole rPPG training signals.

[0184] S723. Divide and denoise each second pixel time-domain change training signal to obtain multiple right sole rPPG training signals.

[0185] S724. Draw multiple dual-sole perfusion training graphs based on the multiple left sole rPPG training signals and the multiple right sole rPPG training signals.

[0186] S725. Generate a target training image based on the multiple dual-sole perfusion training graphs.

[0187] The target training image is a training graph of the change in the ratio of the perfusion index (PI).

[0188] S726. Extract the training mean, training standard deviation, training maximum value, and training minimum value from the target training image as the training features.

[0189] S73. Input the training features into the original classifier for evaluation to obtain a prediction result.

[0190] S74. Calculate a loss function based on the prediction result and the true label corresponding to the target training data.

[0191] S75. Adjust the parameters of the original classifier based on the loss function to obtain the classifier.

[0192] In some embodiments, the dual-sole video evaluation method based on plantar perfusion imaging can also introduce human physiological parameters such as blood oxygen and blood pressure to compare the differences between PAD patients and healthy people, thereby improving the accuracy of PAD grading. At the downsampling level, different sizes of downsampling can be considered to jointly guide the division and drawing of the plantar perfusion map at multiple levels to improve the robustness of the embodiments of the present application.

[0193] Reference Figure 17As shown, it is a schematic block diagram of a dual-foot sole video evaluation device based on plantar perfusion imaging provided in the second aspect of the embodiments of the present application. In Figure 17 the dual-foot sole video evaluation device 100 based on plantar perfusion imaging includes:

[0194] A dual-foot sole video acquisition module 101, configured to acquire a dual-foot sole video of a patient;

[0195] A preprocessing module 102, configured to preprocess the dual-foot sole video to obtain at least one first pixel time-domain change signal and at least one second pixel time-domain change signal;

[0196] A left-foot sole rPPG signal acquisition module 103, configured to perform division and denoising processing on each first pixel time-domain change signal to obtain a plurality of left-foot sole rPPG signals corresponding to each first pixel time-domain change signal;

[0197] A right-foot sole rPPG signal acquisition module 104, configured to perform division and denoising processing on each second pixel time-domain change signal to obtain a plurality of right-foot sole rPPG signals corresponding to each second pixel time-domain change signal;

[0198] A drawing module 105, configured to draw a plurality of dual-foot sole perfusion maps based on the plurality of left-foot sole rPPG signals and the plurality of right-foot sole rPPG signals;

[0199] A generation module 106, configured to generate a target image based on the plurality of dual-foot sole perfusion maps;

[0200] An evaluation module 107, configured to extract the average value, standard deviation, maximum value, and minimum value of the blood perfusion index ratio from the target image and input them into a classifier for evaluation processing to obtain an evaluation result of the dual-foot sole video of the patient.

[0201] Reference Figure 18 As shown, in the third aspect of the embodiments of the present application, a dual-foot sole video evaluation device based on plantar perfusion imaging is further provided. In Figure 18 the dual-foot sole video evaluation device 200 based on plantar perfusion imaging includes: a terminal device 201, an optical imaging device 202, a fixing bracket 203, and a transparent plate 204. The optical imaging device 202 includes a camera 2021 and an LED fill light 2022. Reference Figure 19 is shown;

[0202] The terminal device 201 includes a processor, and the processor is configured to execute the steps of the method for evaluating a dual-foot sole video based on plantar perfusion imaging provided in the first aspect of the embodiments of the present application;

[0203] The transparent plate 204 is placed on the upper part of the fixed bracket 203. During the process of the camera 2021 recording the video of the patient's two plantar surfaces, the patient's two feet are placed on the transparent plate 204, and the LED fill light 2022 irradiates the transparent plate 204;

[0204] The optical imaging device 202 is placed inside the fixed bracket 203;

[0205] After the camera 2021 finishes recording the video of the two plantar surfaces, it sends the video of the two plantar surfaces to the processor.

[0206] The camera 2021 is an RGB camera sensor. Specifically, the camera 2021 can be an RGB camera sensor of the IDS UI-3860-CP model. In other embodiments, the camera 2021 can be any RGB camera sensor including the RGB band of 400 - 700 nm. The camera 2021 is used to obtain a video of continuous image frames including human skin.

[0207] The LED fill light 2022 is an LED light with a continuous broadband spectrum of 400 - 700 nm.

[0208] Reference Figure 19 As shown, the optical imaging device 202, the fixed bracket 203, and the transparent plate 204 form a plantar imager A.

[0209] The fixed bracket 203 is made of steel beams. In other embodiments, other stable anti-torsion materials can be used as the manufacturing material of the fixed bracket 203.

[0210] The transparent plate 204 can be a glass plate or a fully transparent acrylic plate. In other embodiments, other transparent material plates can be placed above the fixed bracket 203 for placing the two feet.

[0211] The embodiment of the present application uses the method of plantar pressurization to enhance the left plantar rPPG signal and the right plantar rPPG signal. Therefore, when the camera records the video, the patient needs to place the plantar surface on the transparent plate 204. This method greatly reduces the noise interference of plantar mechanical movement on the signal, makes the measurement more accurate, generates a more stable plantar perfusion map of the two feet, and can ensure the comfort of the patient during the evaluation process.

[0212] The overall operation of the dual-foot sole video evaluation device 200 based on plantar perfusion imaging is simple and requires less professionalism, making the monitoring of PAD (Peripheral Artery Disease) more convenient and routine. As a result, the evaluation of peripheral artery disease is applicable to home application scenarios or other application scenarios, improving the applicability of PAD evaluation.

[0213] It should also be noted that during the measurement process, the patient only needs to place both feet on the device for 2 minutes, and the evaluation of PAD can be obtained according to the classifier, making the home monitoring of PAD feasible. Based on the current situation of the aging population and the fact that the elderly are a high-risk group for PAD, achieving the goal of home use of PAD evaluation can greatly facilitate the lives of middle-aged and elderly people and increase people's attention to PAD detection.

[0214] The fourth aspect of the embodiments of the present application provides a terminal device. The principle block diagram of the terminal device can be as Figure 20 shown. The terminal device includes a processor, a memory, a network interface, a display screen, and a temperature sensor connected through a system bus. Among them, the processor is used to provide computing and control capabilities. The memory includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the terminal device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, it implements the method for evaluating the dual-foot sole video based on plantar perfusion imaging. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the temperature sensor is pre-set inside the terminal device to detect the operating temperature of the internal device.

[0215] Those skilled in the art can understand that Figure 20 the principle block diagram shown in

[0216] merely shows the block diagram of some structures related to the solution of the present invention, and does not constitute a limitation on the terminal device to which the solution of the present invention is applied. The specific terminal device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0217] In the fifth aspect of the embodiments of the present application, a computer-readable storage medium is provided. The computer-readable storage medium is used to store a computer program, and the computer program causes a computer to execute the steps of the method for evaluating the video of the soles of both feet based on plantar perfusion imaging provided in the first aspect of the embodiments of the present application.

[0218] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the embodiments provided by the present invention can include non-volatile and / or volatile memories. Non-volatile memories can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memories can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0219] Without changing the basic principles of the present application, the technical features of the above embodiments can be combined. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0220] The above embodiments only represent several implementation manners of the present application. Their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the patent application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can be made, and these all belong to the protection scope of the present application. Therefore, the patent protection scope of the present application should be subject to the appended claims.

Claims

1. A method for bilateral sole video assessment based on sole perfusion imaging, characterized in that: include: Obtain videos of the patient's soles; Preprocessing the two-foot sole video to obtain at least one first pixel time-domain change signal and at least one second pixel time-domain change signal; Each first pixel time-domain variation signal is divided and denoised to obtain a plurality of left plantar rPPG signals corresponding to each first pixel time-domain variation signal; Each second pixel time-domain variation signal is divided and denoised to obtain a plurality of right plantar rPPG signals corresponding to each second pixel time-domain variation signal; Drawing a plurality of bipedal plantar perfusion maps based on the plurality of left plantar rPPG signals and the plurality of right plantar rPPG signals; generating a target image based on the plurality of bipedal plantar perfusion maps; Extracting the mean value, standard deviation, maximum value and minimum value of the blood perfusion index ratio from the target image and inputting them into a classifier for evaluation processing to obtain evaluation results of the patient's two-foot sole video; The step of drawing a plurality of bipedal plantar perfusion maps based on the plurality of left plantar rPPG signals and the plurality of right plantar rPPG signals comprises: using a peak detection algorithm to detect the first peak and the first trough of each channel in the target left plantar rPPG signal, wherein the target left plantar rPPG signal is any left plantar rPPG signal among the plurality of left plantar rPPG signals; using the average value of the amplitude difference between the first peak and the first trough of each channel as the first alternating current quantity of the target left plantar rPPG signal; using the average value of the overall amplitude of the target left plantar rPPG signal as the first direct current quantity of the target left plantar rPPG signal; using the first alternating current quantity and the first direct current quantity to calculate the first blood perfusion corresponding to the target left plantar rPPG signal perfusion index; using the peak detection algorithm to detect the second peak and the second trough of each channel in the target right plantar rPPG signal, wherein the target right plantar rPPG signal is any right plantar rPPG signal among the multiple right plantar rPPG signals; using the average value of the amplitude difference between the second peak and the second trough of each channel as the second alternating current quantity of the target right plantar rPPG signal; using the average value of the overall amplitude of the target right plantar rPPG signal as the second direct current quantity of the target right plantar rPPG signal; using the second alternating current quantity and the second direct current quantity to calculate the second blood perfusion index corresponding to the target right plantar rPPG signal; drawing the multiple bilateral plantar perfusion maps based on all the first blood perfusion indices and all the second blood perfusion indices; The steps of generating a target image based on the multiple plantar perfusion maps of both feet include: calculating a first spatial average value of a first blood perfusion index in each plantar perfusion map of both feet; calculating a second spatial average value of a second blood perfusion index in each plantar perfusion map of both feet; connecting all the first spatial average values ​​to form a left foot blood perfusion index change wave; connecting all the second spatial average values ​​to form a right foot blood perfusion index change wave; generating the target image based on the ratio of the left foot blood perfusion index change wave to the right foot blood perfusion index change wave.

2. The method for bilateral sole video assessment based on sole perfusion imaging according to claim 1, characterized in that: The step of preprocessing the two-foot sole video to obtain at least one first pixel time-domain change signal and at least one second pixel time-domain change signal comprises: Decomposing the two-foot sole video to obtain multiple frames of two-foot sole images; Performing average pooling processing on each frame of the plantar image of both feet to obtain an image to be processed corresponding to each frame of the plantar image of both feet; identifying the plantar part of both feet on each image to be processed as a region of interest corresponding to each image to be processed; In each region of interest, a first RGB three-channel pixel value of each pixel point in the left foot region is obtained, and in each region of interest, a second RGB three-channel pixel value of each pixel point in the right foot region is obtained; Based on all first RGB three-channel pixel values, obtaining a first pixel time domain change signal corresponding to each pixel point in the left foot area according to time change; Based on all the second RGB three-channel pixel values, a second pixel time-domain change signal corresponding to each pixel point in the right foot area is obtained according to time change.

3. The method for bilateral sole video assessment based on sole perfusion imaging according to claim 1, characterized in that: The steps of dividing and denoising each first pixel time-domain variation signal to obtain a plurality of left plantar rPPG signals corresponding to each first pixel time-domain variation signal include: Set the sliding window length and step size; Dividing each first pixel time-domain variation signal based on the sliding window length and the step size to obtain a plurality of first sub-signals corresponding to each first pixel time-domain variation signal; A Butterworth filter is used to filter each first sub-signal to obtain a left plantar rPPG signal corresponding to each first sub-signal.

4. The method for bilateral sole video assessment based on sole perfusion imaging according to claim 1, characterized in that: The bipedal sole video assessment method based on sole perfusion imaging further comprises a classifier training step, wherein the classifier training step comprises: Obtain a training data set and a true label set, wherein each training data in the training data set has a one-to-one correspondence with one of the true labels in the true label set; Performing feature extraction processing on target training data to obtain training features, wherein the target training data is any training data in the training data set; Inputting the training features into the original classifier for evaluation processing to obtain a prediction result; Calculate a loss function based on the prediction result and the true label corresponding to the target training data; The parameters of the original classifier are adjusted based on the loss function to obtain the classifier.

5. A two-foot sole video assessment device based on sole perfusion imaging, characterized in that: include: A two-foot sole video acquisition module is used to acquire two-foot sole videos of patients; A preprocessing module, used for preprocessing the two-foot sole video to obtain at least one first pixel time-domain change signal and at least one second pixel time-domain change signal; A left plantar rPPG signal acquisition module, used to divide and denoise each first pixel time-domain variation signal to obtain a plurality of left plantar rPPG signals corresponding to each first pixel time-domain variation signal; A right plantar rPPG signal acquisition module, used for dividing and denoising each second pixel time-domain variation signal to obtain a plurality of right plantar rPPG signals corresponding to each second pixel time-domain variation signal; a drawing module, configured to draw a plurality of bipedal plantar perfusion maps based on the plurality of left plantar rPPG signals and the plurality of right plantar rPPG signals; A generating module, configured to generate a target image based on the plurality of bipedal plantar perfusion maps; An evaluation module, used for extracting the mean value, standard deviation, maximum value and minimum value of the blood perfusion index ratio from the target image and inputting them into a classifier for evaluation processing to obtain an evaluation result of the patient's two-foot sole video; The drawing module is also used to detect the first peak and the first trough of each channel in the target left plantar rPPG signal by using a peak detection algorithm, wherein the target left plantar rPPG signal is any left plantar rPPG signal among the multiple left plantar rPPG signals; the average value of the amplitude difference between the first peak and the first trough of each channel is used as the first alternating current quantity of the target left plantar rPPG signal; the average value of the overall amplitude of the target left plantar rPPG signal is used as the first direct current quantity of the target left plantar rPPG signal; the first alternating current quantity and the first direct current quantity are used to calculate the first blood perfusion index corresponding to the target left plantar rPPG signal; the peak detection algorithm is used to detect the target The second peak and the second trough of each channel in the right plantar rPPG signal, wherein the target right plantar rPPG signal is any one of the multiple right plantar rPPG signals; the average value of the amplitude difference between the second peak and the second trough of each channel is used as the second alternating current of the target right plantar rPPG signal; the average value of the overall amplitude of the target right plantar rPPG signal is used as the second direct current of the target right plantar rPPG signal; the second alternating current and the second direct current are used to calculate the second blood perfusion index corresponding to the target right plantar rPPG signal; and the multiple plantar perfusion maps of both feet are drawn based on all the first blood perfusion indices and all the second blood perfusion indices; The generation module is also used to calculate the first spatial average value of the first blood perfusion index in each plantar perfusion map of both feet; calculate the second spatial average value of the second blood perfusion index in each plantar perfusion map of both feet; connect all the first spatial average values ​​to form a left foot blood perfusion index change wave; connect all the second spatial average values ​​to form a right foot blood perfusion index change wave; and generate the target image based on the ratio of the left foot blood perfusion index change wave and the right foot blood perfusion index change wave.

6. A two-foot sole video assessment device based on sole perfusion imaging, characterized in that: include: Terminal equipment, optical imaging equipment, a fixing bracket and a transparent plate, wherein the optical imaging equipment includes a camera and an LED fill light; The terminal device comprises a processor, and the processor is used to execute the steps of the bipedal plantar video assessment method based on plantar perfusion imaging as described in any one of claims 1 to 4 above; The transparent plate is placed on the upper part of the fixed bracket. When the camera is recording the video of the soles of the patient's feet, the patient's feet are placed on the transparent plate, and the LED fill light is irradiated onto the transparent plate. The optical imaging device is placed inside the fixed bracket; After recording the video of the soles of both feet, the camera sends the video of the soles of both feet to the processor.

7. A terminal device, characterized in that: include: A processor and a memory, the memory being used to store a computer program, the processor being used to call and run the computer program stored in the memory to execute the steps of the bipedal plantar video assessment method based on plantar perfusion imaging as described in any one of claims 1 to 4.

8. A computer-readable storage medium, characterized in that: Used to store a computer program, wherein the computer program enables a computer to execute the steps of the bipedal plantar video assessment method based on plantar perfusion imaging as described in any one of claims 1 to 4.

Images