A method and system for analyzing nailfold blood microcirculation image data

Abstract

The present invention relates to the field of image data analysis technology, and specifically discloses a method and system for analyzing nailfold blood microcirculation image data. The method comprises: collecting raw image data of nailfold blood microcirculation, preprocessing the raw image data of nailfold blood microcirculation, segmenting nailfold blood vessels using a deep learning-based U-Net convolutional neural network model, calculating nailfold blood flow velocity using a particle image velocimetry algorithm, correcting the nailfold blood flow velocity, constructing a nailfold blood vessel-blood flow-loop joint model and a nailfold blood microcirculation analysis influence model, updating a nailfold blood joint analysis factor, and using a support vector machine model to output a classification result of the nailfold blood microcirculation image data. The present invention solves the problems of low accuracy and efficiency of traditional nailfold blood microcirculation image data analysis, insufficient consideration of interference factors in data analysis, and insufficient analysis accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of image data analysis, and in particular to a nailfold blood microcirculation image data analysis method and system. Background Art

[0002] The nailfold is rich in microvascular tissue. Advances in optical imaging technology have led to the emergence of high-resolution, high-definition nailfold microcirculatory imaging equipment, enabling the acquisition of clearer and richer microcirculatory image information. Simultaneously, advancements in computer technology and digital image processing algorithms, including image denoising, segmentation, and feature extraction, have provided the technical means for in-depth analysis of nailfold microcirculatory images. Analysis of nailfold microcirculatory imaging data can help explore the mechanisms of microcirculatory changes and provide a theoretical basis for new therapeutic targets and intervention strategies.

[0003] Currently, research on nailfold microcirculation imaging data analysis still has some shortcomings. Specifically, the traditional method of visually observing nailfold microcirculation images is inaccurate and inefficient, failing to meet the requirements for precise analysis of nailfold microcirculation imaging data. Interference factors in nailfold microcirculation imaging data analysis are not fully considered, resulting in insufficient analysis accuracy. The accuracy and reliability of the results still need to be improved. Summary of the Invention

[0004] In view of the deficiencies in the prior art, the present invention provides a nailfold blood microcirculation image data analysis method and system, which can effectively solve the problems involved in the above-mentioned background technology.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: In the first aspect, the present invention provides a method for analyzing nailfold blood microcirculation image data, comprising the following steps: using a microcirculation imaging device to collect original image data of nailfold blood microcirculation, and preprocessing the original image data of nailfold blood microcirculation; using a U-Net convolutional neural network model based on deep learning to segment nailfold blood vessels; using a particle image velocimetry algorithm to calculate the nailfold blood flow velocity in the segmented nailfold blood vessel image, and correcting the nailfold blood flow velocity; collecting nailfold blood vessel shape; and The nailfold blood state characteristic data, nailfold blood flow characteristic data and nailfold blood loop environment characteristic data are used to construct a nailfold blood vessel-blood flow-loop joint model, and output the nailfold blood joint analysis factor; the nailfold blood microcirculation analysis characteristic data are obtained, the nailfold blood microcirculation analysis influence model is constructed, and the nailfold blood microcirculation analysis influence factor is output; the nailfold blood joint analysis factor is corrected based on the nailfold blood microcirculation analysis influence factor to obtain the updated nailfold blood joint analysis factor; based on the updated nailfold blood joint analysis factor, the support vector machine model is used to output the nailfold blood microcirculation imaging data classification result.

[0006] As a further method, a microcirculation imaging device is used to collect raw image data of the nailfold blood microcirculation, and the raw image data of the nailfold blood microcirculation is preprocessed. The specific analysis process is as follows: the raw image data of the nailfold blood microcirculation is collected using the microcirculation imaging device; the raw image data of the nailfold blood microcirculation is denoised using the non-local mean denoising algorithm:

[0007] ;

[0008] in, is the value of pixel p after denoising, is the normalization constant, is the set of all pixels in the original image data of nail fold blood microcirculation, is the weight between pixels p and q, is the grayscale value of the original image at pixel q;

[0009] ;

[0010] Among them, i traverses the various position offsets in the neighborhood window centered on p and q, With standard deviation The weighted Euclidean distance of , h is the parameter that controls the denoising strength, is the grayscale value of the original image at position p+i in the neighborhood window centered on pixel p, is the grayscale value of the original image at position q+i in the neighborhood window centered on pixel q;

[0011] Adopting the adaptive grayscale correction method based on histogram equalization, the grayscale histogram of the image is adjusted to a uniform distribution to enhance the contrast of the image: for a grayscale image, the pixel value after histogram equalization is Calculated by the following formula:

[0012] ;

[0013] Where L is the total number of gray levels, is the size of the image, is the frequency of occurrence of gray level a, is the rounding function, is the grayscale value of the pixel at the coordinate (x, y) in the image; the preprocessed nailfold blood microcirculation image data is obtained.

[0014] As a further approach, a U-Net convolutional neural network model based on deep learning is used to segment nail fold blood vessels. The specific analysis process is as follows: a large number of labeled nail fold blood microcirculation images are collected as a training set, the U-Net convolutional neural network model based on deep learning is trained, and the cross-entropy loss function is used to measure the difference between the predicted results and the true labels:

[0015] ;

[0016] Where b is the sample label, b=1,2,3,..., , is the number of samples, c is the category label, c=1,2,3,...,C, C is the number of categories, is the true label of sample b belonging to category c, is the probability that the model predicts that sample b belongs to category c;

[0017] The model parameters are continuously adjusted through the back-propagation algorithm to minimize the cross-entropy loss function. The pre-processed nailfold blood microcirculation image data is input into the U-Net convolutional neural network model based on deep learning. The U-Net convolutional neural network model based on deep learning outputs the probability that each pixel belongs to a nailfold blood vessel. Based on the nailfold blood vessel probability division rules stored in the database, the segmented nailfold blood vessel image is obtained.

[0018] The probability division rule of nail fold blood vessels is:

[0019] ;

[0020] in, The U-Net convolutional neural network model based on deep learning outputs the probability that the pixel belongs to the nail fold blood vessel. The threshold value for the nailfold blood vessel probability stored in the database is set.

[0021] As a further method, the nail fold blood vessel image is segmented and the nail fold blood flow velocity is calculated using the particle image velocimetry algorithm, and the nail fold blood flow velocity is corrected. The specific analysis process is as follows: in the segmented nail fold blood vessel image, based on the window division scheme stored in the database, the nail fold blood vessel images of two adjacent frames are divided into two parts. and Divide into multiple small windows and calculate the displacement of pixels in each window to estimate blood flow velocity:

[0022] ;

[0023] in, and is the pixel grayscale value at the coordinate (x, y) in the nailfold blood vessel images of two adjacent frames, for The mean grayscale value within the window, for The mean grayscale value within the window, is the normalized cross-correlation function, is the displacement of pixels in the image;

[0024] The normalized cross-correlation function The largest The actual displacement Blood flow velocity The actual displacement and frame interval Calculation yields:

[0025] ;

[0026] The angle between the blood vessel and the image plane is , correct the nail fold blood flow velocity to obtain the corrected nail fold blood flow velocity :

[0027] .

[0028] As a further method, a nailfold blood vessel-blood flow-circumference joint model is constructed to output the nailfold blood joint analysis factor. The specific analysis process is: based on the nailfold blood vessel segmentation image and the nailfold blood flow velocity, the nailfold blood vessel morphological characteristic data, nailfold blood flow characteristic data and nailfold blood loop environment characteristic data are collected, among which: the nailfold blood vessel morphological characteristic data specifically includes the nailfold blood vessel input branch diameter , nailfold blood vessel output branch diameter , nailfold blood vessel length ; The nail fold blood flow characteristic data specifically includes nail fold blood flow velocity , nailfold red blood cell aggregation area density , nailfold white blood cell count ; Nailfold blood loop environmental characteristics data specifically include the size of the subpapillary venous plexus , perivascular bleeding area ; Construct a nailfold blood vessel-blood flow-peripheral joint model, and output the nailfold blood joint analysis factor based on the nailfold blood vessel morphological characteristic data, nailfold blood flow characteristic data and nailfold blood periphery environmental characteristic data. The nailfold blood joint analysis factor serves as the analysis basis for obtaining the updated nailfold blood joint analysis factor.

[0029] As a further method, the nailfold blood vessel-blood flow-peripheral joint model has the following specific analysis process:

[0030] ;

[0031] Where, It is a nailfold blood combined analysis factor. It is a factor for analyzing nail fold blood vessels. It is the nailfold blood flow analysis factor, It is the factor for analyzing the nail fold blood loop environment. For the setting The weight factor, For the setting The weight factor, For the setting The weight factor of , e is a natural constant.

[0032] As a further method, a nail fold blood microcirculation analysis influence model is constructed to output the nail fold blood microcirculation analysis influence factors. The specific analysis process is: obtaining the nail fold blood microcirculation analysis characteristic data, which specifically includes the nail fold blood microcirculation detection environment temperature , resolution of nailfold blood microcirculation detection imaging equipment , nailfold blood microcirculation detection imaging equipment magnification ; Construct a nailfold blood microcirculation analysis impact model, and output the nailfold blood microcirculation analysis impact factor based on the nailfold blood microcirculation analysis characteristic data. The nailfold blood microcirculation analysis impact factor serves as the analysis basis for obtaining the updated nailfold blood joint analysis factor;

[0033] The nail fold blood microcirculation analysis impact model, the specific analysis process is as follows:

[0034] ;

[0035] Where, is the influencing factor of nailfold blood microcirculation analysis, and e is the natural constant.

[0036] As a further method, the nailfold-blood combined analysis factor was modified based on the nailfold blood microcirculation analysis influencing factor to obtain an updated nailfold-blood combined analysis factor. The specific analysis process is as follows:

[0037] ;

[0038] Where, To update the nailfold blood combined analysis factor, It is a nailfold blood combined analysis factor. For the setting The weight factor, For the setting The weight factor of .

[0039] As a further method, based on the updated nailfold-blood joint analysis factor, the support vector machine model is used to output the classification results of nailfold-blood microcirculation imaging data. The specific analysis process is as follows: the nailfold-blood sample data of known status stored in the database is obtained and divided into training set and test set according to the set ratio; the support vector machine model is trained using the training set, and the objective function is minimized. , while satisfying the constraints and ;in, is the Lagrange multiplier corresponding to the i-th nailfold blood sample, is the Lagrange multiplier corresponding to the j-th nailfold blood sample, is the i-th nailfold blood sample, is the jth nailfold blood sample, i and j are the nailfold blood sample numbers, n is the total number of nailfold blood samples, C is the penalty parameter, is the radial basis kernel function, , is the width parameter of the radial basis kernel function, is the feature vector of the i-th nailfold blood sample, is the characteristic vector of the j-th nailfold blood sample; the trained support vector machine model is input into the preprocessed updated nailfold blood joint analysis factor, the state of the nailfold blood is classified through the decision function, and the classification result of the nailfold blood microcirculation image data is output.

[0040] The second aspect of the present invention provides a nail fold blood microcirculation image data analysis system, comprising an original image data preprocessing module, a blood vessel segmentation module, a blood flow velocity analysis and correction module, a joint analysis factor output module, an analysis influence factor output module, a joint analysis factor update module and a classification result output module, wherein: the original image data preprocessing module is used to use a microcirculation imaging device to collect nail fold blood microcirculation original image data and preprocess the nail fold blood microcirculation original image data; the blood vessel segmentation module is used to use a U-Net convolutional neural network model based on deep learning to perform nail fold blood vessel segmentation; the blood flow velocity analysis and correction module is used to calculate the nail fold blood flow velocity in the segmented nail fold blood vessel image using a particle image velocimetry algorithm, and perform blood flow analysis on the nail fold blood flow velocity. The degree of correction is performed; the joint analysis factor output module is used to collect nail fold blood vessel morphological characteristic data, nail fold blood flow characteristic data and nail fold blood loop environment characteristic data, build a nail fold blood vessel-blood flow-loop joint model, and output the nail fold blood joint analysis factor; the analysis influence factor output module is used to obtain nail fold blood microcirculation analysis characteristic data, build a nail fold blood microcirculation analysis influence model, and output the nail fold blood microcirculation analysis influence factor; the joint analysis factor update module is used to correct the nail fold blood joint analysis factor based on the nail fold blood microcirculation analysis influence factor to obtain an updated nail fold blood joint analysis factor; the classification result output module is used to output the classification result of the nail fold blood microcirculation image data based on the updated nail fold blood joint analysis factor using a support vector machine model.

[0041] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects:

[0042] (1) The present invention provides a method and system for analyzing nailfold blood microcirculation image data. The U-Net convolutional neural network model can accurately segment nailfold blood vessels. Compared with traditional methods, it can better identify complex vascular structures, provide an accurate basis for subsequent blood flow velocity calculation and feature data collection, and reduce diagnostic errors caused by inaccurate vascular segmentation. The method collects multiple feature data such as nailfold blood vessel morphology, blood flow, and perivascular environment, constructs a joint model, and outputs joint analysis factors. At the same time, the joint analysis factors are corrected by considering the influencing factors of microcirculation analysis. This analysis method that integrates multi-dimensional information can more comprehensively reflect the true status of nailfold blood microcirculation.

[0043] (2) The present invention provides a tool for studying the interrelationships among various microcirculatory factors by constructing a nailfold blood vessel-blood flow-circular joint model and a nailfold blood microcirculation analysis influence model. Through continuous optimization and verification of these models, the understanding of the nailfold blood microcirculation mechanism can be further improved. A variety of advanced technologies and algorithms such as deep learning, particle image velocimetry, and support vector machines are comprehensively utilized to improve the accuracy and reliability of data processing. The gradual correction process can reduce the accumulation of errors and improve the accuracy of the final classification results. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The present invention is further described with reference to the accompanying drawings. However, the embodiments in the accompanying drawings do not constitute any limitation to the present invention. A person skilled in the art can obtain other drawings based on the following drawings without creative effort.

[0045] Figure 1 Schematic diagram of the method steps of the present invention.

[0046] Figure 2 This is a schematic diagram of system module connections of the present invention. DETAILED DESCRIPTION

[0047] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0048] Reference Figure 1 As shown, the first aspect of the present invention provides a method for analyzing nail fold blood microcirculation image data, including: using a microcirculation imaging device to collect nail fold blood microcirculation original image data, and preprocessing the nail fold blood microcirculation original image data.

[0049] The specific analysis process is as follows: using microcirculation imaging equipment to collect raw image data of nailfold blood microcirculation; using non-local means denoising algorithm to denoise the raw image data of nailfold blood microcirculation:

[0050] ;

[0051] in, is the value of pixel p after denoising, is the normalization constant, is the set of all pixels in the original image data of nail fold blood microcirculation, is the weight between pixels p and q, is the grayscale value of the original image at pixel q;

[0052] ;

[0053] Among them, i traverses the various position offsets in the neighborhood window centered on p and q, With standard deviation The weighted Euclidean distance of , h is the parameter that controls the denoising strength, is the grayscale value of the original image at position p+i in the neighborhood window centered on pixel p, is the grayscale value of the original image at position q+i in the neighborhood window centered on pixel q;

[0054] Adopting the adaptive grayscale correction method based on histogram equalization, the grayscale histogram of the image is adjusted to a uniform distribution to enhance the contrast of the image: for a grayscale image, the pixel value after histogram equalization is Calculated by the following formula:

[0055] ;

[0056] Where L is the total number of gray levels, is the size of the image, is the frequency of occurrence of gray level a, is the rounding function, is the grayscale value of the pixel at the coordinate (x, y) in the image; the preprocessed nailfold blood microcirculation image data is obtained.

[0057] The non-local means denoising algorithm leverages image self-similarity to effectively remove noise from the original image while preserving key structural information, such as blood vessels. This reduces noise interference with subsequent analysis and improves image clarity. Adaptive grayscale correction using histogram equalization reduces grayscale deviations caused by uneven light sources, enhancing image contrast and making it easier to distinguish blood vessels from surrounding tissue, facilitating subsequent segmentation and feature extraction.

[0058] High-quality preprocessed images provide a reliable foundation for subsequent vessel segmentation and blood flow velocity calculation. Denoising and grayscale correction reduce data errors, enabling the deep learning-based U-Net model to more accurately identify vessels and the particle image velocimetry algorithm to more accurately calculate blood flow velocity, ultimately improving the accuracy of the entire analysis process. Preprocessed data is more standardized and clear, facilitating model construction and further analysis.

[0059] The U-Net convolutional neural network model based on deep learning is used to segment nail fold blood vessels.

[0060] The specific analysis process is as follows: a large number of annotated nailfold blood microcirculation images are collected as a training set, a U-Net convolutional neural network model based on deep learning is trained, and the cross-entropy loss function is used to measure the difference between the predicted results and the actual labels:

[0061] ;

[0062] Where b is the sample label, b=1,2,3,..., , is the number of samples, c is the class label, c=1,2,3,...,C, C is the number of categories (in this implementation, C=2, i.e., vascular and non-vascular), is the true label of sample b belonging to category c (in this embodiment, the true label of sample b belonging to category c is 0 or 1), is the probability that the model predicts that sample b belongs to category c;

[0063] The model parameters are continuously adjusted through the back-propagation algorithm to minimize the cross-entropy loss function. The pre-processed nailfold blood microcirculation image data is input into the U-Net convolutional neural network model based on deep learning. The U-Net convolutional neural network model based on deep learning outputs the probability that each pixel belongs to a nailfold blood vessel. Based on the nailfold blood vessel probability division rules stored in the database, the segmented nailfold blood vessel image is obtained.

[0064] The probability division rule of nail fold blood vessels is:

[0065] ;

[0066] in, The U-Net convolutional neural network model based on deep learning outputs the probability that the pixel belongs to the nail fold blood vessel. The threshold value for the nailfold blood vessel probability stored in the database is set.

[0067] The U-Net model's unique structure, combining contraction and expansion paths, effectively extracts image features and restores resolution, enabling high-precision segmentation of complex structures such as nailfold blood vessels. By training on a large number of annotated images, using a cross-entropy loss function to measure differences and backpropagating parameters, the model accurately learns vascular features and outputs the probability that each pixel belongs to a vessel, resulting in precisely segmented images and reducing the errors and subjectivity of manual segmentation.

[0068] There is no need for manual pixel-by-pixel segmentation. Simply inputting the pre-processed image data into the model can automatically output the segmentation results, greatly saving manpower and time costs and improving work efficiency. It is suitable for processing large amounts of nailfold blood microcirculation image data and meeting the needs of rapid data processing in clinical and research.

[0069] As long as there is sufficiently diverse and accurately labeled training data, the model can adapt to nailfold blood microcirculation images of different individuals and under different imaging conditions. It can also adjust model parameters or combine other technologies for further optimization based on new research or clinical needs to improve segmentation effects and expand the scope of application.

[0070] In the segmented nailfold blood vessel image, the nailfold blood flow velocity is calculated using the particle image velocimetry algorithm, and the nailfold blood flow velocity is corrected.

[0071] The specific analysis process is as follows: in the segmented nail fold blood vessel image, based on the window division scheme stored in the database, the two adjacent frames of nail fold blood vessel images are divided into and Divide into multiple small windows and calculate the displacement of pixels in each window to estimate blood flow velocity:

[0072] ;

[0073] in, and is the pixel grayscale value at the coordinate (x, y) in the nailfold blood vessel images of two adjacent frames, for The mean grayscale value within the window, for The mean grayscale value within the window, is the normalized cross-correlation function, is the displacement of pixels in the image;

[0074] The normalized cross-correlation function The largest The actual displacement Blood flow velocity The actual displacement and frame interval Calculation yields:

[0075] ;

[0076] The angle between the blood vessel and the image plane is , correct the nail fold blood flow velocity to obtain the corrected nail fold blood flow velocity :

[0077] .

[0078] By dividing adjacent frames of nailfold blood vessel images into small windows and calculating pixel displacement within those windows to estimate blood flow velocity, the method captures detailed intravascular blood flow. The application of a normalized cross-correlation function accurately determines the actual pixel displacement, resulting in more precise measurements and providing reliable blood flow velocity data for analyzing nailfold microcirculation.

[0079] The effect of the angle between the blood vessel and the image plane on blood flow velocity measurement is taken into account and corrected, compensating for measurement errors caused by imaging angle. This makes the measurement results closer to the actual blood flow velocity, enhances the reliability and scientific nature of the data, and helps to more accurately assess microcirculatory function.

[0080] This method simulates the actual flow of blood in blood vessels and analyzes displacement at the pixel level, which is consistent with the physiological characteristics of nailfold blood microcirculation. This is of great significance for in-depth research on the mechanism of nailfold blood microcirculation.

[0081] The nailfold blood vessel morphological characteristic data, nailfold blood flow characteristic data and nailfold blood loop environment characteristic data are collected to construct a nailfold blood vessel-blood flow-loop joint model and output the nailfold blood joint analysis factor.

[0082] The specific analysis process is as follows: based on the nail fold blood vessel segmentation image and the nail fold blood flow velocity, the nail fold blood vessel morphological characteristic data, nail fold blood flow characteristic data and nail fold blood loop surrounding environment characteristic data are collected, among which: the nail fold blood vessel morphological characteristic data specifically includes the nail fold blood vessel input branch diameter , nailfold blood vessel output branch diameter , nailfold blood vessel length ; The nail fold blood flow characteristic data specifically includes nail fold blood flow velocity , nailfold red blood cell aggregation area density , nailfold white blood cell count ; Nailfold blood loop environmental characteristics data specifically include the size of the subpapillary venous plexus , perivascular bleeding area ; Construct a nailfold blood vessel-blood flow-peripheral joint model, and output the nailfold blood joint analysis factor based on the nailfold blood vessel morphological characteristic data, nailfold blood flow characteristic data and nailfold blood periphery environmental characteristic data. The nailfold blood joint analysis factor serves as the analysis basis for obtaining the updated nailfold blood joint analysis factor.

[0083] The nailfold blood vessel-blood flow-peripheral joint model has the following specific analysis process:

[0084] ;

[0085] Where, It is a nailfold blood combined analysis factor. It is a factor for analyzing nail fold blood vessels. It is the nailfold blood flow analysis factor, It is the factor for analyzing the nail fold blood loop environment. For the setting The weight factor, For the setting The weight factor, For the setting The weight factor of , e is a natural constant.

[0086] The system collects multi-dimensional data covering vascular morphology, blood flow characteristics, and the peri-loop environment, breaking through the limitations of single-index analysis. For example, it considers not only the diameter and length of the blood vessels, but also blood flow velocity, blood cell aggregation, the size of the peri-loop venous plexus, and the bleeding area. This allows for a comprehensive and holistic reflection of the true state of nailfold blood microcirculation.

[0087] By integrating various characteristic data into joint analysis factors through a specific formula, the microcirculatory status can be quantitatively represented. Compared with qualitative descriptions, quantitative indicators are more precise and objective. The combined analysis factors in the joint model are combined through weighting factors to reflect the interrelationships between vascular, blood flow, and perivascular environmental factors. This better reflects the actual conditions of the nailfold blood microcirculatory system and can more accurately assess the synergistic effects of various factors on microcirculation.

[0088] Obtain the characteristic data of nailfold blood microcirculation analysis, construct the nailfold blood microcirculation analysis impact model, and output the nailfold blood microcirculation analysis impact factor.

[0089] The specific analysis process is: obtaining the nail fold blood microcirculation analysis characteristic data, which specifically includes the nail fold blood microcirculation detection environment temperature , resolution of nailfold blood microcirculation detection imaging equipment , nailfold blood microcirculation detection imaging equipment magnification ; Construct a nailfold blood microcirculation analysis impact model, and output the nailfold blood microcirculation analysis impact factor based on the nailfold blood microcirculation analysis characteristic data. The nailfold blood microcirculation analysis impact factor serves as the analysis basis for obtaining the updated nailfold blood joint analysis factor;

[0090] The nail fold blood microcirculation analysis impact model, the specific analysis process is as follows:

[0091] ;

[0092] Where, To analyze the influencing factors of nailfold blood microcirculation.

[0093] External factors such as the detection environment temperature, imaging equipment resolution and magnification are included in the analysis, recognizing that these factors will affect the nailfold blood microcirculation imaging data and analysis results, avoiding analysis deviations caused by ignoring these factors, and making the analysis more comprehensive and objective.

[0094] External factors are quantified into influencing factors of nailfold blood microcirculation analysis through specific formulas, providing a specific numerical reference for evaluating the degree of influence of external factors on microcirculation analysis. This serves as the basis for updating the nailfold blood joint analysis factor. This influencing factor can correct the joint analysis factor, reduce the error caused by external factors, and make the final analysis results closer to the actual nailfold blood microcirculation status, thereby improving the accuracy and reliability of nailfold blood microcirculation imaging data analysis.

[0095] The nailfold-blood combined analysis factor was modified based on the influencing factor of nailfold blood microcirculation analysis to obtain the updated nailfold-blood combined analysis factor.

[0096] The specific analysis process is as follows:

[0097] ;

[0098] Where, To update the nailfold blood combined analysis factor, It is a nailfold blood combined analysis factor. For the setting The weight factor, For the setting The weight factor of .

[0099] The nailfold microcirculation analysis influencing factor integrates external factors such as the test environment temperature and imaging equipment parameters. By using this factor to modify the joint analysis factor, the interference and bias caused by these external factors can be eliminated or reduced, allowing the updated joint analysis factor to more accurately reflect the true state of nailfold microcirculation.

[0100] Different testing environments and equipment conditions can produce data discrepancies. This correction mechanism allows analysis results to adapt to diverse testing scenarios. Regardless of the environment in which data is collected, the adjustment of influencing factors ensures comparability and stability of the combined analysis factors, improving the applicability and versatility of the entire analysis method.

[0101] By combining external influencing factors with the inherent characteristics of microcirculation, the analysis system for nailfold blood microcirculation has been improved. This has expanded from simply analyzing the internal characteristics of microcirculation to also taking into account external influencing factors, forming a more comprehensive and systematic analysis process that helps to further explore relevant information about nailfold blood microcirculation.

[0102] Based on the updated nailfold-blood joint analysis factors, the support vector machine model is used to output the classification results of nailfold-blood microcirculation imaging data.

[0103] The specific analysis process is as follows: obtain the nailfold blood sample data of known state stored in the database, divide it into training set and test set according to the set ratio; use the training set to train the support vector machine model, and minimize the objective function , while satisfying the constraints and ;in, is the Lagrange multiplier corresponding to the i-th nailfold blood sample, is the Lagrange multiplier corresponding to the j-th nailfold blood sample, is the i-th nailfold blood sample, is the jth nailfold blood sample, i and j are the nailfold blood sample numbers, n is the total number of nailfold blood samples, C is the penalty parameter, is the radial basis kernel function, , is the width parameter of the radial basis kernel function, is the feature vector of the i-th nailfold blood sample, is the characteristic vector of the j-th nailfold blood sample; the trained support vector machine model is input into the preprocessed updated nailfold blood joint analysis factor, the state of the nailfold blood is classified through the decision function, and the classification result of the nailfold blood microcirculation image data is output.

[0104] The support vector machine model effectively handles both linear and nonlinear classification problems by finding the optimal classification hyperplane in high-dimensional space. Training with nailfold blood sample data with known states can learn the mapping relationship between sample characteristics and states. The updated nailfold-blood joint analysis factor comprehensively considers multiple factors, providing the model with more representative and accurate data. This enables the model to more accurately classify nailfold blood microcirculation imaging data, improving diagnostic accuracy.

[0105] During training, the support vector machine model effectively avoids overfitting by minimizing the objective function and satisfying the constraints, resulting in good generalization. This means the model not only performs well on the training data but also provides reliable classification results for new, unseen nailfold blood samples, making it suitable for assessing the nailfold blood microcirculatory status of different individuals.

[0106] This method leverages the updated nailfold blood joint analysis factors obtained through previous processing. These factors integrate information about nailfold blood vessels, blood flow, the periarticular environment, and external influencing factors. The support vector machine model learns and classifies this rich data, unlocking the hidden insights behind the data and providing a more comprehensive analysis of the state of nailfold blood microcirculation.

[0107] Reference Figure 2 As shown, the second aspect of the present invention provides a nailfold blood microcirculation image data analysis system, including an original image data preprocessing module, a blood vessel segmentation module, a blood flow velocity analysis and correction module, a joint analysis factor output module, an analysis influence factor output module, a joint analysis factor update module and a classification result output module.

[0108] The raw image data preprocessing module is used to use the microcirculation imaging device to collect the raw image data of the nail fold blood microcirculation and preprocess the raw image data of the nail fold blood microcirculation.

[0109] The blood vessel segmentation module is used to segment nail fold blood vessels using the U-Net convolutional neural network model based on deep learning.

[0110] The blood flow velocity analysis and correction module is used to calculate the nail fold blood flow velocity in the segmented nail fold blood vessel image using the particle image velocimetry algorithm and to correct the nail fold blood flow velocity.

[0111] The joint analysis factor output module is used to collect nail fold blood vessel morphological characteristic data, nail fold blood flow characteristic data and nail fold blood loop environment characteristic data, build a nail fold blood vessel-blood flow-loop joint model, and output nail fold blood joint analysis factors.

[0112] The analysis impact factor output module is used to obtain the nail fold blood microcirculation analysis characteristic data, build the nail fold blood microcirculation analysis impact model, and output the nail fold blood microcirculation analysis impact factor.

[0113] The joint analysis factor updating module is used to modify the nailfold-blood joint analysis factor based on the nailfold-blood microcirculation analysis influencing factor to obtain an updated nailfold-blood joint analysis factor.

[0114] The classification result output module is used to output the classification results of nailfold blood microcirculation image data based on the updated nailfold blood joint analysis factor and the support vector machine model.

[0115] The above content is merely an example and explanation of the structure of the present invention. Those skilled in the art may make various modifications or additions to the described specific embodiments or replace them in a similar manner. As long as they do not deviate from the structure of the invention or exceed the scope defined by the claims, they should all fall within the scope of protection of the present invention.

Claims

1. A nailfold blood microcirculation image data analysis method, characterized in that: The following steps are involved: Using microcirculation imaging equipment, collecting the original image data of the nail fold blood microcirculation, and preprocessing the original image data of the nail fold blood microcirculation; The U-Net convolutional neural network model based on deep learning is used to segment nail fold blood vessels; In the segmented nailfold blood vessel image, the nailfold blood flow velocity is calculated using the particle image velocimetry algorithm and the nailfold blood flow velocity is corrected; Collect nailfold blood vessel morphological characteristic data, nailfold blood flow characteristic data, and nailfold blood loop environment characteristic data, build a nailfold blood vessel-blood flow-loop joint model, and output nailfold blood joint analysis factors; Obtain nailfold blood microcirculation analysis characteristic data, construct nailfold blood microcirculation analysis influence model, and output nailfold blood microcirculation analysis influence factors; Based on the influencing factors of nailfold blood microcirculation analysis, the nailfold blood joint analysis factors were revised to obtain the updated nailfold blood joint analysis factors; Based on the updated nailfold-blood joint analysis factors, the support vector machine model is used to output the classification results of nailfold-blood microcirculation imaging data; The nailfold blood vessel-blood flow-circumference joint model is constructed to output the nailfold blood joint analysis factor. The specific analysis process is as follows: According to the nail fold blood vessel segmentation image and the nail fold blood flow velocity, the nail fold blood vessel morphological characteristic data, the nail fold blood flow characteristic data and the nail fold blood loop surrounding environment characteristic data are collected, wherein: The nail fold blood vessel morphological characteristic data specifically includes the nail fold blood vessel input branch diameter , nailfold blood vessel output branch diameter , nailfold blood vessel length ; Nail fold blood flow characteristic data specifically includes nail fold blood flow velocity , nailfold red blood cell aggregation area density , nailfold white blood cell count ; Nailfold blood loop environmental characteristics include the size of the subpapillary venous plexus , perivascular bleeding area ; Constructing a nailfold blood vessel-blood flow-peripheral joint model, based on nailfold blood vessel morphological characteristic data, nailfold blood flow characteristic data, and nailfold blood peripheral environmental characteristic data, outputting a nailfold blood joint analysis factor, which serves as the analysis basis for obtaining an updated nailfold blood joint analysis factor; The specific analysis process of the nailfold blood vessel-blood flow-peripheral joint model is as follows: ； Where, It is a nailfold blood combined analysis factor. It is a factor for analyzing nail fold blood vessels. It is the nailfold blood flow analysis factor, It is the factor for analyzing the nail fold blood loop environment. For the setting The weight factor, For the setting The weight factor, For the setting The weight factor of , e is a natural constant.

2. The method for analyzing nailfold blood microcirculation image data according to claim 1, characterized in that: The microcirculation imaging device is used to collect the original image data of the nail fold blood microcirculation and pre-process the original image data of the nail fold blood microcirculation. The specific analysis process is as follows: Use microcirculation imaging equipment to collect raw image data of nailfold blood microcirculation; The non-local means denoising algorithm is used to denoise the raw image data of the nail fold blood microcirculation: ； in, is the value of pixel p after denoising, is the normalization constant, is the set of all pixels in the original image data of nail fold blood microcirculation, is the weight between pixels p and q, is the grayscale value of the original image at pixel q; ； Among them, i traverses the various position offsets in the neighborhood window centered on p and q, With standard deviation The weighted Euclidean distance of , h is the parameter that controls the denoising strength, is the grayscale value of the original image at position p+i in the neighborhood window centered on pixel p, is the grayscale value of the original image at position q+i in the neighborhood window centered on pixel q; Adopting the adaptive grayscale correction method based on histogram equalization, the grayscale histogram of the image is adjusted to a uniform distribution to enhance the contrast of the image: For a grayscale image, the pixel value after histogram equalization Calculated by the following formula: ； Where L is the total number of gray levels, is the size of the image, is the frequency of occurrence of gray level a, is the rounding function, is the grayscale value of the pixel at the coordinate (x, y) in the image; The preprocessed nailfold blood microcirculation image data is obtained.

3. The method for analyzing nailfold blood microcirculation image data according to claim 1, characterized in that: The U-Net convolutional neural network model based on deep learning is used to segment the nail fold blood vessels. The specific analysis process is as follows: A large number of annotated nail fold blood microcirculation images were collected as a training set, and the U-Net convolutional neural network model based on deep learning was trained. The cross entropy loss function was used to measure the difference between the predicted results and the actual labels: ； Where b is the sample label, b=1,2,3,..., , is the number of samples, c is the category label, c=1,2,3,...,C, C is the number of categories, is the true label of sample b belonging to category c, is the probability that the model predicts that sample b belongs to category c; The model parameters are continuously adjusted through the back-propagation algorithm to minimize the cross entropy loss function; The preprocessed nailfold blood microcirculation image data is input into a U-Net convolutional neural network model based on deep learning. The U-Net convolutional neural network model based on deep learning outputs the probability that each pixel belongs to a nailfold blood vessel. Based on the nailfold blood vessel probability segmentation rules stored in the database, a segmented nailfold blood vessel image is obtained. The probability division rule of nail fold blood vessels is: ； in, The U-Net convolutional neural network model based on deep learning outputs the probability that the pixel belongs to the nail fold blood vessel. The threshold value for the nailfold blood vessel probability stored in the database is set.

4. The method for analyzing nailfold blood microcirculation image data according to claim 1, wherein: In the segmented nail fold blood vessel image, the nail fold blood flow velocity is calculated using the particle image velocimetry algorithm, and the nail fold blood flow velocity is corrected. The specific analysis process is as follows: In the segmented nail fold blood vessel images, based on the window division scheme stored in the database, the two adjacent frames of nail fold blood vessel images are segmented. and Divide into multiple small windows and calculate the displacement of pixels in each window to estimate blood flow velocity: ； in, and is the pixel grayscale value at the coordinate (x, y) in the nailfold blood vessel images of two adjacent frames, for The mean grayscale value within the window, for The mean grayscale value within the window, is the normalized cross-correlation function, is the displacement of pixels in the image; The normalized cross-correlation function The largest Recorded as actual displacement ; Blood flow velocity The actual displacement and frame interval Calculation yields: ； The angle between the blood vessel and the image plane is , correct the nail fold blood flow velocity to obtain the corrected nail fold blood flow velocity : 。 5. The method for analyzing nailfold blood microcirculation image data according to claim 1, characterized in that: The nailfold blood microcirculation analysis influencing model is constructed and the nailfold blood microcirculation analysis influencing factors are output. The specific analysis process is as follows: Obtain nail fold blood microcirculation analysis characteristic data, which specifically includes nail fold blood microcirculation detection environment temperature , resolution of nailfold blood microcirculation detection imaging equipment , nail fold blood microcirculation detection imaging equipment magnification ; Construct a nailfold blood microcirculation analysis impact model, and output the nailfold blood microcirculation analysis impact factor based on the nailfold blood microcirculation analysis feature data. The nailfold blood microcirculation analysis impact factor serves as the analysis basis for obtaining the updated nailfold blood joint analysis factor. The nail fold blood microcirculation analysis impact model, the specific analysis process is as follows: ； Where, is the influencing factor of nailfold blood microcirculation analysis, and e is the natural constant.

6. The method for analyzing nailfold blood microcirculation image data according to claim 4, characterized in that: The nailfold-blood combined analysis factor is corrected based on the nailfold-blood microcirculation analysis influencing factor to obtain an updated nailfold-blood combined analysis factor. The specific analysis process is as follows: ； Where, To update the nailfold blood combined analysis factor, It is a nailfold blood combined analysis factor. For the setting The weight factor, For the setting The weight factor of .

7. The method for analyzing nailfold blood microcirculation image data according to claim 1, characterized in that: Based on the updated nailfold-blood joint analysis factor, the support vector machine model is used to output the classification results of the nailfold-blood microcirculation image data. The specific analysis process is as follows: Obtain nailfold blood sample data of known status stored in the database and divide it into training set and test set according to the set ratio; The support vector machine model is trained using the training set by minimizing the objective function , while satisfying the constraints and ; in, is the Lagrange multiplier corresponding to the i-th nailfold blood sample, is the Lagrange multiplier corresponding to the j-th nailfold blood sample, is the i-th nailfold blood sample, is the jth nailfold blood sample, i and j are the nailfold blood sample numbers, n is the total number of nailfold blood samples, C is the penalty parameter, is the radial basis kernel function, , is the width parameter of the radial basis kernel function, is the feature vector of the i-th nailfold blood sample, is the j-th nailfold blood sample feature vector; The trained support vector machine model is input into the preprocessed updated nailfold blood joint analysis factor, the nailfold blood status is classified through the decision function, and the classification results of the nailfold blood microcirculation image data are output.

8. A nailfold blood microcirculation image data analysis system, applied to a nailfold blood microcirculation image data analysis method according to any one of claims 1 to 6, characterized in that: It includes raw image data preprocessing module, blood vessel segmentation module, blood flow velocity analysis and correction module, joint analysis factor output module, analysis influence factor output module, joint analysis factor update module and classification result output module, among which: The raw image data preprocessing module is used to collect the raw image data of the nail fold blood microcirculation using a microcirculation imaging device and preprocess the raw image data of the nail fold blood microcirculation; The blood vessel segmentation module is used to segment nail fold blood vessels using a U-Net convolutional neural network model based on deep learning; The blood flow velocity analysis and correction module is used to calculate the nail fold blood flow velocity in the segmented nail fold blood vessel image using a particle image velocimetry algorithm and to correct the nail fold blood flow velocity; The joint analysis factor output module is used to collect nail fold blood vessel morphological characteristic data, nail fold blood flow characteristic data and nail fold blood loop environment characteristic data, build a nail fold blood vessel-blood flow-loop joint model, and output nail fold blood joint analysis factors; The analysis influence factor output module is used to obtain nail fold blood microcirculation analysis characteristic data, construct a nail fold blood microcirculation analysis influence model, and output the nail fold blood microcirculation analysis influence factor; The joint analysis factor updating module is used to modify the nailfold-blood joint analysis factor based on the nailfold-blood microcirculation analysis influencing factor to obtain an updated nailfold-blood joint analysis factor; The classification result output module is used to output the classification result of the nailfold blood microcirculation image data based on the updated nailfold blood joint analysis factor and using the support vector machine model; The nailfold blood vessel-blood flow-circumference joint model is constructed to output the nailfold blood joint analysis factor. The specific analysis process is as follows: According to the nail fold blood vessel segmentation image and the nail fold blood flow velocity, the nail fold blood vessel morphological characteristic data, the nail fold blood flow characteristic data and the nail fold blood loop surrounding environment characteristic data are collected, wherein: The nail fold blood vessel morphological characteristic data specifically includes the nail fold blood vessel input branch diameter , nailfold blood vessel output branch diameter , nailfold blood vessel length ; Nail fold blood flow characteristic data specifically includes nail fold blood flow velocity , nailfold red blood cell aggregation area density , nailfold white blood cell count ; Nailfold blood loop environmental characteristics include the size of the subpapillary venous plexus , perivascular bleeding area ; Constructing a nailfold blood vessel-blood flow-peripheral joint model, based on nailfold blood vessel morphological characteristic data, nailfold blood flow characteristic data, and nailfold blood peripheral environmental characteristic data, outputting a nailfold blood joint analysis factor, which serves as the analysis basis for obtaining an updated nailfold blood joint analysis factor; The specific analysis process of the nailfold blood vessel-blood flow-peripheral joint model is as follows: ； Where, It is a nailfold blood combined analysis factor. It is a factor for analyzing nail fold blood vessels. It is the nailfold blood flow analysis factor, It is the factor for analyzing the nail fold blood loop environment. For the setting The weight factor, For the setting The weight factor, For the setting The weight factor of , e is a natural constant.

Images