A method for analyzing muscle ultrasound images of small model animals

Abstract

The present invention discloses a method for analyzing muscle ultrasound images of model small animals, the method comprising: dividing the muscle ultrasound images of the model small animals into ROIs, and extracting muscle morphological features, image frequency features, and image texture features from the divided muscle ultrasound images; performing statistical analysis on the features to obtain statistically different features, and performing repeatability analysis based on test-retest reliability to obtain reliable features; according to the statistically different features and reliable features, obtaining an analysis model of the muscle ultrasound images; inputting the muscle ultrasound images of the model small animals to be analyzed into the analysis model of the muscle ultrasound images for analysis, and obtaining the detection and quantitative evaluation results of the muscle structure of the model small animals to be analyzed. The present invention can collect ultrasound images of the muscle parts of model small animals in vivo, and has the advantages of being non-invasive, simple and easy to implement, and can achieve fine detection and visualization of the muscle structure of living model small animals.

Description

Technical Field

[0001] The present invention relates to the field of medical image analysis, and particularly to a method for analyzing muscle ultrasound images of model small animals. Background Art

[0002] Research has found that the gene sequences of various model small animals are highly homologous to those of humans, and their functional responses are similar to those of humans. Therefore, through small animal models, the occurrence and development laws of human diseases can be explored more effectively. Secondly, by using model small animals, a large number of exploration experiments can be carried out to obtain more experimental data, which helps to make the evaluation results more accurate, scientific, and reliable. Among them, detecting the muscles of model small animals can provide valuable information for the evaluation of muscle function and medical analysis of various diseases.

[0003] Currently, the commonly used detection methods for the muscles of model small animals include histological detection, skeletal muscle contraction property detection, limb grip strength measurement, pole climbing test, rotarod test, hanging test, open field test, etc. The above detection methods are either invasive or can only detect the muscle strength and function of model small animals, and cannot achieve fine detection and quantitative evaluation of the muscle structure of living model small animals.

[0004] Therefore, the existing technology still needs to be improved and developed. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide an image feature extraction and classification method based on muscle ultrasound aiming at the above-mentioned defects of the existing technology, aiming to solve the problem that the fine detection and quantitative evaluation of the muscle structure of living model small animals cannot be achieved in the existing technology.

[0006] The technical solution adopted by the present invention to solve the technical problem is as follows:

[0007] In the first aspect, the present invention provides a method for analyzing muscle ultrasound images of model small animals, wherein the method includes:

[0008] Obtain the muscle ultrasound image of the model small animal, and perform ROI division on the muscle ultrasound image to obtain the divided muscle ultrasound image;

[0009] Extract muscle morphological features, image frequency features, and image texture features from the divided muscle ultrasound image;

[0010] Perform statistical analysis on the muscle morphological features, image frequency features, and image texture features to obtain statistically different features, and perform repeatability analysis based on test-retest reliability to obtain reliable features;

[0011] An analysis model of the muscle ultrasound image is obtained according to the statistically different features and the reliable features;

[0012] The muscle ultrasound image of the model small animal to be analyzed is input into the analysis model of the muscle ultrasound image for analysis, and the detection and quantitative evaluation results of the muscle structure of the model small animal to be analyzed are obtained.

[0013] In one implementation, the obtaining of the muscle ultrasound image of the model small animal includes:

[0014] Depilate the hind limbs of the test model small animal. Under the anesthesia state of the test model small animal, place the test model small animal supine in the center of the experimental plate, and place its mouth and nose in the pipeline connected to the anesthesia instrument;

[0015] Set the detection mode of the ultrasonic imaging system to the musculoskeletal detection mode;

[0016] Align the long axis of the ultrasonic probe with the Achilles tendon of the test model small animal, and keep the ultrasonic probe placed at the first detection position on the hind limb of the test model small animal by setting marks;

[0017] Under the anesthesia state of the test model small animal, apply an ultrasonic gel couplant to ensure the acoustic coupling between the ultrasonic probe and the skin, adjust the ultrasonic probe to optimize the contrast of the muscle bundles in the ultrasonic image, and use a real-time B-mode ultrasonic imaging device to obtain the muscle ultrasound image of the test model small animal based on the detection mode and the first detection position.

[0018] In one implementation, the performing ROI division on the muscle ultrasound image to obtain the divided muscle ultrasound image includes:

[0019] Perform grayscale conversion, cropping on the muscle ultrasound image, and divide the ROI according to the manually outlined or automatically segmented muscle region to obtain the divided muscle ultrasound image.

[0020] In one implementation, the extracting of the muscle morphological features and image frequency features from the divided muscle ultrasound image includes:

[0021] Perform normalized Radon transform on the divided muscle ultrasound image to obtain a Radon transform matrix;

[0022] Calculate the gradient of the Radon transform matrix and perform edge enhancement to obtain a Radon transform gradient matrix;

[0023] Obtain the muscle thickness, muscle fiber length and pennation angle features according to the Radon transform gradient matrix;

[0024] Based on the muscle thickness, muscle fiber length, and pennation angle characteristics, the muscle morphological characteristics are obtained;

[0025] Extract the average frequency analysis feature from the segmented muscle ultrasound image; wherein, the calculation formula of the average frequency analysis feature is n, I, and f are the length, power, and frequency of the power density spectrum, respectively.

[0026] In one implementation, extracting the image texture analysis feature from the segmented muscle ultrasound image includes:

[0027] Based on the calculation of pixel gray-scale distribution, extract the first-order statistical features from the segmented muscle ultrasound image; wherein, the first-order statistical features include integrated optical density, mean value, standard deviation, variance, skewness, kurtosis, and energy;

[0028] Based on the calculation of the gray-level co-occurrence matrix, extract the Haralick features from the segmented muscle ultrasound image; wherein, the Haralick features include contrast, correlation, energy, entropy, homogeneity, and symmetry;

[0029] Based on the calculation of the gray-level run length matrix, extract the Galloway features from the segmented muscle ultrasound image; wherein, the Galloway features include short run dominance, long run dominance, gray-level non-uniformity, long run non-uniformity, and run percentage;

[0030] Based on the comparison result between the central pixel of the local region of the image and the neighborhood of the central pixel, extract the local binary pattern features from the segmented muscle ultrasound image; wherein, the local binary pattern features include energy and entropy, and the energy is LBP energy =∑ i f i 2 , and the entropy is LBP entropy =-∑ i f i 2 log 2 (f i ), f i represents the corresponding frequency of the i-th block in the local region of the image;

[0031] Based on the first-order statistical features, the Haralick features, the Galloway features, and the local binary pattern features, the image texture analysis features are obtained.

[0032] In one implementation, performing statistical analysis on the muscle morphological characteristics, image frequency characteristics, and image texture characteristics to obtain statistically different characteristics includes:

[0033] Perform a normality test on the muscle morphological features, image frequency features, and image texture features to obtain a first test result;

[0034] If the first test result follows a normal distribution, perform a homogeneity of variance test on the first test result to obtain a second test result;

[0035] If the first test result does not follow a normal distribution, perform a non-parametric test on the muscle morphological features, image frequency features, and image texture features to obtain the statistically different features;

[0036] If the second test result satisfies the homogeneity of variance, perform an independent samples T-test on the muscle morphological features, image frequency features, and image texture features to obtain the statistically different features;

[0037] If the second test result does not satisfy the homogeneity of variance, perform a modified variance T-test on the muscle morphological features, image frequency features, and image texture features to obtain the statistically different features.

[0038] In one implementation, the performing of the repeatability analysis based on test-retest reliability to obtain reliable features includes:

[0039] Perform a repeatability test on the muscle morphological features, image frequency features, and image texture features to obtain a number of test results;

[0040] Evaluate the reliability of the number of test results using the intraclass correlation coefficient ICC(1,1); where the correlation coefficient is where BMS and WMS are the root mean squares between groups and within groups obtained by Kruskal-Wallis one-way analysis of variance, respectively, and k is the number of repeated measurements;

[0041] Obtain the reliable features according to the reliability.

[0042] In a second aspect, an embodiment of the present invention further provides a muscle ultrasound image analysis device for model small animals, where the device includes:

[0043] An ROI division module, configured to obtain a muscle ultrasound image of a model small animal and perform ROI division on the muscle ultrasound image to obtain a divided muscle ultrasound image;

[0044] A feature extraction module, configured to extract muscle morphological features, image frequency features, and image texture features from the divided muscle ultrasound image;

[0045] A feature analysis module, configured to perform statistical analysis on the muscle morphological features, image frequency features, and image texture features to obtain statistically different features, and perform repeatability analysis based on test-retest reliability to obtain reliable features;

[0046] An analysis model acquisition module, configured to obtain an analysis model of the muscle ultrasound image according to the statistically different features and the reliable features;

[0047] An analysis module, configured to input the muscle ultrasound image of the model small animal to be analyzed into the analysis model of the muscle ultrasound image for analysis, and obtain the detection and quantitative evaluation results of the muscle structure of the model small animal to be analyzed.

[0048] In a third aspect, an embodiment of the present invention further provides an intelligent terminal, where the intelligent terminal includes a memory, a processor, and a muscle ultrasound image analysis program for model small animals stored in the memory and executable on the processor. When the processor executes the muscle ultrasound image analysis program for model small animals, the steps of the muscle ultrasound image analysis method for model small animals as described in any one of the above are implemented.

[0049] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, where a muscle ultrasound image analysis program for model small animals is stored on the computer-readable storage medium. When the muscle ultrasound image analysis program for model small animals is executed by a processor, the steps of the muscle ultrasound image analysis method for model small animals as described in any one of the above are implemented.

[0050] Beneficial effects: Compared with the prior art, the present invention provides a method for analyzing muscle ultrasound images of model small animals. First, the present invention uses the currently widely used ultrasound equipment to collect ultrasound images of the muscle parts of model small animals in vivo, which has the advantages of non-invasive, simple and easy to implement, and can achieve fine detection and visualization of the muscle structure of model small animals in vivo. Then, image processing and other means are used to extract muscle morphological features, image frequency analysis features, image texture features, etc. from the obtained muscle ultrasound images of model small animals, so as to realize the quantitative evaluation of the muscle structure of model small animals, and use statistical and repeatability analysis methods to analyze the extracted image features, so as to screen out important features highly related to diseases and stable and reliable features with high test-retest reliability, and construct an analysis system for muscle ultrasound images of model small animals. Finally, by inputting the muscle ultrasound image of the model small animal to be analyzed, the corresponding features can be extracted and the analysis results can be obtained, so as to realize the fine detection and quantitative evaluation of the muscle structure of model small animals in vivo. Description of the Drawings

[0051] To more clearly illustrate the technical solutions in the embodiments of the present invention or 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 recorded in the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0052] Figure 1 It is a schematic flowchart of a method for analyzing muscle ultrasound images of model small animals provided by an embodiment of the present invention.

[0053] Figure 2 It is an exemplary diagram of a muscle ultrasound image of the hind limb of a model small animal provided by an embodiment of the present invention.

[0054] Figure 3 It is a flowchart of statistical analysis provided by an embodiment of the present invention.

[0055] Figure 4 A principle block diagram of a device for analyzing muscle ultrasound images of model small animals provided by an embodiment of the present invention.

[0056] Figure 5 It is a principle block diagram of the internal structure of an intelligent terminal provided by an embodiment of the present invention. Detailed implementation manners

[0057] To make the objectives, technical solutions and effects of the present invention clearer and more definite, the following further elaborates on the present invention with reference to the accompanying drawings and by way of examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0058] Studies have found that the gene sequences of various model small animals are highly homologous to those of humans, and their functional responses are similar to those of humans. Therefore, through small animal models, the occurrence and development laws of human diseases can be explored more effectively. Secondly, by using model small animals, a large number of exploration experiments can be carried out to obtain more experimental data, which helps to make the evaluation results more accurate, scientific and reliable, and is conducive to detecting early lesions in organisms and improving the early diagnosis ability of diseases. The application of model small animals has a significant promoting effect on the research of the pathogenesis of human diseases, the analysis of the disease development process, the research of new drugs and new treatment methods, the evaluation of treatment effects, the research and development of treatment instruments, and the prevention and treatment of diseases. Therefore, model small animals are considered to be crucial experimental methods and means in modern life science research, and in vivo model small animal imaging technology plays an irreplaceable role in medical research. In vivo model small animal imaging technology refers to the non-destructive detection of the structure, function and physiological information of small animals without damaging them, which can realize the long-term dynamic observation of small animals. For small animal imaging with various human disease models established, through long-term continuous monitoring, the occurrence, development and various treatment effects of diseases can be studied. In addition, the research on live small animals is carried out in the real in vivo environment of the animals, and all the interacting physiological information is integrated during the research process, which can make the experimental results more real and reliable. Detecting the muscles of model small animals can provide valuable information for the muscle function evaluation and medical diagnosis of various diseases, and can also provide valuable information for monitoring the muscle response during the disease progression and treatment process, which is of great significance for the muscle function evaluation, disease diagnosis and rehabilitation plan formulation of various diseases.

[0059] Currently, there are mainly seven existing detection methods for the muscles of model small animals: the first is histological detection based on the physiological and biochemical indexes of muscle tissue; the second is the detection of skeletal muscle contraction characteristics based on muscle contractility; the third is the determination of the claw grip force of the four limbs based on the grip strength; the fourth is the pole climbing test based on the climbing time and state; the fifth is the rotarod test based on the running time of the rotarod; the sixth is the hanging test based on the endurance of the four limbs; the seventh is the open field test based on the total autonomous crawling path.

[0060] Histological examination refers to taking samples from the muscles of model small animals and measuring their tissue biochemical indices, evaluating histopathological indices, and measuring molecular biological indices. Measuring tissue biochemical indices mainly involves measuring muscle protein content, muscle glycogen, muscle creatine kinase, etc.; evaluating histopathological indices refers to observing and evaluating the histological characteristics of muscles through conventional histology (such as hematoxylin-eosin staining), special histology (such as Masson staining, myosin ATPase histochemical staining, modified Gomori staining, etc.), and ultrastructural histology (transmission electron microscopy); measuring molecular biological indices refers to detecting and analyzing skeletal muscle mechanogrowth factor, actin, and myosin heavy chain through semi-quantitative reverse transcription polymerase chain reaction with an internal control. Although histological examination can meticulously detect various physiological and biochemical indices of muscles and can quantitatively analyze and evaluate the morphological structure of muscles, this detection method is invasive and requires sacrificing small animals for sampling. This leads to the need for a large number of small animal individuals and also means that long-term monitoring and follow-up observations cannot be conducted on the same small animal individual.

[0061] Detection of skeletal muscle contraction characteristics refers to perfusing isolated muscles such as soleus muscle and extensor digitorum longus muscle at 30 °C and applying a certain voltage stimulus to measure their single twitch force and tetanic contraction force. During the incubation of the muscle, after applying the optimal voltage stimulus to the muscle, the maximum contraction force at the optimal initial length of the muscle is measured, which is the single twitch force; then a tetanic stimulus is applied at the optimal voltage of the muscle, and the tetanic contraction force of the muscle at each stimulation time point is recorded in real time to obtain the corresponding contraction force curve. Although the detection of skeletal muscle contraction characteristics can detect the contraction characteristics of muscles, this detection method is also invasive and requires sacrificing small animals for sampling. This leads to the need for a large number of small animal individuals and also means that long-term monitoring and follow-up observations cannot be conducted on the same small animal individual.

[0062] Determination of the claw grip strength of the four limbs refers to measuring the grasping strength of small animals through a claw grip strength tester, which can evaluate the strength of the forelimbs and hindlimbs of small animals, thereby evaluating the degree of nerve injury and muscle injury in the four limbs of small animals. Place the small animal on the metal grid measurement area of the claw grip strength tester. The small animal will automatically grasp the metal grid. After it grasps firmly, pull its tail backward. The small animal will instinctively grasp the metal grid to resist. At this time, the tester will generate a reading. After pulling it away from the metal grid, record the maximum reading of the tester during this period, which is the strength of the forelimbs and hindlimbs of this small animal. Usually, each small animal is measured three times and the average value is taken. Although this detection method can measure the grasping strength of the four limbs of model small animals, it cannot achieve fine detection and quantitative evaluation of the muscle structure of living model small animals.

[0063] The pole climbing test refers to evaluating the climbing speed, limb coordination ability, and the strength of the four limbs' support of small animals by testing the climbing time and state of small animals. The climbing time of small animals is recorded through a pole climbing device and a timer. Before the formal test, the small animals are trained first to avoid situations where they stop moving forward or climb upward in the opposite direction. During the formal test, the small animals are placed at the top of the pole and the timing starts simultaneously. The timing stops when their heads touch the ground, and the recorded time is the climbing time. If a small animal shows a situation where its muscle strength is insufficient to support it to complete the entire pole climbing experiment, it is classified into single-leg weakness, double-leg weakness, and four-limbed weakness according to the degree. Although this detection method can measure the climbing speed, limb coordination ability, and the strength of the four limbs' support of model small animals, it cannot achieve fine detection and quantitative evaluation of the muscle structure of living model small animals.

[0064] The rotarod test refers to evaluating the balance ability, coordination ability, and muscle strength of small animals by testing the running time of model small animals on a continuously rotating rod. Before the experiment, the small animals are first subjected to rotarod adaptation and learning training for five consecutive days. When the rotarod test starts, the rotation speed of the rotarod instrument is adjusted to an appropriate speed, and the specific speed can be selected according to the body size of each model small animal, etc. Record the time of the small animal on the rotarod instrument. Taking a certain time as the cut-off value, if it exceeds this time, it is recorded according to this time, and if it is less than this time, it is recorded according to the actual time. Repeat three times, and take a certain rest time between each rotarod test. Although this detection method can evaluate the balance ability, coordination ability, and muscle strength of model small animals, it cannot achieve fine detection and quantitative evaluation of the muscle structure of living model small animals.

[0065] The hanging test refers to measuring the limb endurance of each model small animal through a hanging net. The small animal is placed upside down on the hanging grid. After it grabs firmly, start timing. Stop timing when it falls from the hanging grid due to exhaustion of strength. Taking a certain time as the cut-off value, if it exceeds this time, it is recorded according to this time, and if it is less than this time, it is recorded according to the actual time. Repeat three times, and give the small animal a certain interval time to recover its physical strength each time. Although this detection method can measure the limb endurance of model small animals, it cannot achieve fine detection and quantitative evaluation of the muscle structure of living model small animals.

[0066] The open field test refers to evaluating the locomotor autonomy and motor ability of small animals by counting the total self-propelled crawling path of small animals in an open box within a certain period. The small animal is placed in a light-dark box, and its movement trajectory within the box within a certain period is tracked and recorded by a high-definition camera. The activity trajectory map is analyzed using open field test video analysis software, and the total crawling path is counted. Although this detection method can evaluate the locomotor autonomy and motor ability of model small animals, it cannot achieve fine detection and quantitative evaluation of the muscle structure of living model small animals.

[0067] Therefore, in view of the above problems, the present invention provides a method for analyzing muscle ultrasound images of model small animals. First, the present invention uses currently widely used ultrasound equipment to collect ultrasound images of the muscle parts of model small animals in vivo, which has the advantages of non-invasive, simple and easy to implement, and can achieve fine detection and visualization of the muscle structure of living model small animals. Then, by means of image processing and other means, morphological features, image frequency analysis features, image texture features, etc. of the muscle ultrasound images of the obtained model small animals are extracted, so as to realize the quantitative evaluation of the muscle structure of the model small animals. And statistical and repeatability analysis methods are used to analyze the extracted image features, so as to screen out important features highly related to diseases and stable and reliable features with high test-retest reliability, and construct a muscle ultrasound image analysis system for model small animals. Finally, by inputting the muscle ultrasound images of the model small animals to be analyzed, the corresponding features can be extracted and the analysis results can be obtained, so as to realize the fine detection and quantitative evaluation of the muscle structure of living model small animals.

[0068] Exemplary method

[0069] This embodiment provides a method for analyzing muscle ultrasound images of model small animals. As Figure 1 shown, the method includes the following steps:

[0070] Step S100, obtain a muscle ultrasound image of a model small animal, and perform ROI division on the muscle ultrasound image to obtain a divided muscle ultrasound image;

[0071] Specifically, ultrasound imaging equipment has the potential to detect the fine structure of muscles, allows visualization and quantification of muscle structures, and has the advantages of non-invasiveness, no radiation, real-time dynamic monitoring, etc. After obtaining the muscle ultrasound image of the model small animal in this embodiment, it is necessary to perform ROI division on the muscle ultrasound image to obtain a divided muscle ultrasound image.

[0072] Among them, ROI (region of interest) is the region of interest. In machine vision and image processing, the region that needs to be processed outlined in the form of a square, circle, ellipse, irregular polygon, etc. from the processed image is called the region of interest ROI. Various operators and functions are commonly used on machine vision software such as Halcon, OpenCV, and Matlab to obtain the region of interest ROI and perform the next step of image processing.

[0073] In one implementation, step S100 of this embodiment includes the following steps:

[0074] Step S101: Depilate the hind limbs of the small animal model under test. Place the small animal model under test in a supine position at the center of the experimental board in its anesthetized state, and place its mouth and nose in the tube connected to the anesthetic apparatus;

[0075] Step S102: Set the detection mode of the ultrasonic imaging system to the musculoskeletal detection mode;

[0076] Step S103: Align the long axis of the ultrasonic probe with the Achilles tendon of the small animal model under test, and keep the ultrasonic probe placed at the first detection position on the hind limb of the small animal model under test by setting marks;

[0077] Step S104: Under the anesthetized state of the small animal model under test, apply an ultrasonic gel couplant to ensure acoustic coupling between the ultrasonic probe and the skin, adjust the ultrasonic probe to optimize the contrast of muscle bundles in the ultrasonic image, and use a real-time B-mode ultrasonic imaging device to obtain the muscle ultrasonic image of the small animal model under test based on the detection mode and the first detection position;

[0078] Step S105: Grayscale, crop the muscle ultrasonic image, and divide the ROI according to the manually outlined or automatically segmented muscle area to obtain the divided muscle ultrasonic image.

[0079] Specifically, the specific acquisition steps for the muscle ultrasonic image of the hind limb of the small animal model are as follows: First, before the ultrasonic image acquisition, depilate the hind limb of the small animal model to avoid the influence of its fur on the acquisition and quality of the muscle ultrasonic image; Second, place the small animal model in a supine position at the center of the experimental board in its anesthetized state, and place its mouth and nose in the tube connected to the anesthetic apparatus to keep it in an anesthetized state during the experiment, which is convenient for the acquisition of ultrasonic images; Then use a real-time ultrasonic imaging device to obtain the muscle ultrasonic image of its hind limb, where the ultrasonic imaging system selects the musculoskeletal detection mode, align the long axis of the ultrasonic probe with the Achilles tendon, and place it on the surface of the hind limb or other specific positions; Apply an appropriate amount of ultrasonic gel couplant to ensure acoustic coupling between the probe and the skin; The ultrasonic probe can be adjusted to optimize the contrast of muscle bundles in the ultrasonic image, and marks can be made on the position to ensure that the probe is placed at the same position each time. An example of the muscle ultrasonic image of the hind limb of the small animal model is Figure 2 as shown. And preprocess the image before feature extraction, including grayscaling, cropping, and dividing the ROI according to the manually outlined or automatically segmented muscle area.

[0080] It should be noted that in the acquisition of muscle ultrasound images, in addition to acquiring muscle ultrasound images of model small animals in a static state, muscle ultrasound images during the structural dynamic changes caused by muscle stretching of model small animals can also be acquired; in addition to B-mode ultrasound equipment, shear wave elastography equipment or other ultrasound imaging methods can also be used for acquisition.

[0081] Step S200: Extract muscle morphological features, image frequency features, and image texture features from the segmented muscle ultrasound images;

[0082] In one implementation, step S200 of this embodiment includes the following steps:

[0083] Step S201: Perform a normalized Radon transform on the segmented muscle ultrasound images to obtain a Radon transform matrix;

[0084] Step S202: Calculate the gradient of the Radon transform matrix and perform edge enhancement to obtain a Radon transform gradient matrix;

[0085] Step S203: Obtain muscle thickness, muscle fiber length, and pennation angle features according to the Radon transform gradient matrix;

[0086] Step S204: Obtain the muscle morphological features according to the muscle thickness, muscle fiber length, and pennation angle features;

[0087] Step S205: Extract the average frequency analysis features from the segmented muscle ultrasound images; where the calculation formula for the average frequency analysis features is n, I, and f are the length, power, and frequency of the power density spectrum respectively;

[0088] Step S206: Extract first-order statistical features from the segmented muscle ultrasound images based on pixel gray-level distribution calculation; where the first-order statistical features include integrated optical density, mean value, standard deviation, variance, skewness, kurtosis, and energy;

[0089] Step S207: Extract Haralick features from the segmented muscle ultrasound images based on gray-level co-occurrence matrix calculation; where the Haralick features include contrast, correlation, energy, entropy, homogeneity, and symmetry;

[0090] Step S208: Extract Galloway features from the segmented muscle ultrasound images based on gray-level run length matrix calculation; where the Galloway features include short run dominance, long run dominance, gray-level inhomogeneity, long run inhomogeneity, and run percentage;

[0091] Step S209: Extract local binary pattern features from the segmented muscle ultrasound image based on the comparison result between the central pixel of the local region of the image and the neighborhood of the central pixel. Among them, the local binary pattern features include energy and entropy, and the energy is LBP energy = ∑ i f i 2 , and the entropy is LBP entropy = -∑ i f i 2 log 2 (f i ), where f i represents the corresponding frequency of the i-th block in the local region of the image;

[0092] Step S210: Obtain the image texture analysis features according to the first-order statistical features, the Haralick features, the Galloway features, and the local binary pattern features.

[0093] Specifically, the morphological features of muscles include muscle thickness, muscle fiber length, and pennation angle, etc. The length and pennation angle of muscle fibers are both calculated based on muscle fibers. However, since the hindlimb muscles of some model small animals such as mice are small, the muscle fibers in the muscle region cannot be well displayed. Therefore, the morphological features of their hindlimb muscles are mainly the area of the muscle region outlined manually or automatically segmented by algorithms. If the muscle fibers in the collected muscle region are relatively clear, some feature detection methods can also be used to estimate the morphological parameters of muscles, such as the feature detection method based on the Radon transform gradient matrix, etc.

[0094] The mean frequency analysis feature (MFAF), as an effective parameter related to muscle mass and expected to describe the structural differences of skeletal muscles, is not significantly affected by the configurations of different ultrasound devices. The calculation formula is as follows:

[0095]

[0096] Among them, n, I, and f are the length, power, and frequency of the power density spectrum respectively.

[0097] Image texture analysis features mainly include first-order statistical features and high-order texture features. On the one hand, according to previous studies, first-order statistical features can effectively and quantitatively describe the ultrasonic echo intensity of skeletal muscle; in addition, there are also studies showing that there are differences in the ultrasonic echo intensity information of skeletal muscle in different ages or groups, and these features can also provide some structural information related to muscle state, thus providing effective information for muscle injury assessment. On the other hand, high-order texture features, such as Haralick features, Galloway features, and Local Binary Pattern (LBP) features, etc., can perform better than first-order statistical features in fine tasks such as muscle gender recognition. Specifically, first-order statistical features are features directly calculated based on the pixel gray distribution of the original image, including integrated optical density, mean, standard deviation, variance, skewness, kurtosis, and energy, etc. Haralick features are calculated from the gray-level co-occurrence matrix, which is a matrix function of pixel distance and direction, calculating the correlation between the gray values of two points at a given spatial distance d and direction θ, and is a common method to describe image texture by studying the spatial correlation characteristics of gray levels. Typical Haralick features include contrast, correlation, energy, entropy, homogeneity, and symmetry, etc. Usually, each Haralick feature contains four directions: 0°, 45°, 90°, and 135°. Galloway features are calculated based on the gray-level run-length matrix, which represents the regularity of texture changes in an image, and its size is determined by the gray level of the image and the image size, including texture statistical features such as short-run dominance, long-run dominance, gray-level non-uniformity, long-run non-uniformity, and run percentage, etc., and each Galloway feature also contains four directions: 0°, 45°, 90°, and 135°. LBP features are obtained by comparing the central pixel of a local region of the image with its neighborhood, mainly describing the local texture features of the image, and having significant advantages such as rotation invariance and gray-level invariance, including energy and entropy, and the specific calculation formulas are as follows:

[0098] LBP energy = ∑ i f i 2

[0099] LBP entropy = -∑ i f i 2 log 2 (f i )

[0100] where f i represents the corresponding frequency of the i-th block in the local region of the image.

[0101] It should be noted that when extracting muscle morphological features from muscle ultrasound images, such as pennation angle, muscle thickness, muscle fiber angle, fascicle length, physiological cross-sectional area of muscle, etc., in addition to using the feature detection method based on the Radon transform gradient matrix, methods based on Hough transform, deep learning, etc. can also be used to locate the structural elements of muscle tissue, so as to realize the automatic measurement of morphological parameters.

[0102] Step S300: Conduct statistical analysis on the muscle morphological features, image frequency features, and image texture features to obtain statistically different features, and conduct repeatability analysis based on test-retest reliability to obtain reliable features;

[0103] Statistical analysis is to collect a large amount of data / numbers, and use statistical methods to find regular features from a large amount of data, so as to visualize and make the large amount of data intuitive, in order to understand the statistical laws contained in the data, and then use the regular features of the data to explain problems. Repeatability analysis is to evaluate the test-retest reliability level of image features through multiple repeated measurements to screen out stable and reliable features.

[0104] Specifically, in this embodiment, statistical and repeatability analysis methods are used to analyze the extracted image features, so as to screen out important features highly related to diseases and stable and reliable features with high test-retest reliability.

[0105] In one implementation, step S300 of this embodiment includes the following steps:

[0106] Step S301: Conduct a normality test on the muscle morphological features, image frequency features, and image texture features to obtain the first test result;

[0107] Step S302: If the first test result follows a normal distribution, conduct a homogeneity of variance test on the first test result to obtain the second test result;

[0108] Step S303: If the first test result does not follow a normal distribution, conduct a non-parametric test on the muscle morphological features, image frequency features, and image texture features to obtain the statistically different features;

[0109] Step S304: If the second test result satisfies the homogeneity of variance, conduct an independent samples T-test on the muscle morphological features, image frequency features, and image texture features to obtain the statistically different features;

[0110] Step S305: If the second test result satisfies non - homogeneity of variance, perform a modified variance T - test on the muscle morphological features, image frequency features, and image texture features to obtain the statistically significant features;

[0111] Step S306: Perform a repeatability test on the muscle morphological features, image frequency features, and image texture features to obtain a number of test results;

[0112] Step S307: Evaluate the reliability using the intra - class correlation coefficient ICC(1,1) for the number of test results; where the correlation coefficient is where BMS and WMS are the root mean squares between groups and within groups obtained from Kruskal - Wallis one - way analysis of variance respectively, and k is the number of repeated measurements;

[0113] Step S308: Obtain the reliable features based on the reliability.

[0114] Specifically, perform statistical analysis and repeatability analysis on the extracted image features. The specific steps of the analysis are as follows: First, screen out the features with significant differences between groups through statistical analysis. When performing statistical analysis, first perform a normality test and a homogeneity of variance test. If the data follows a normal distribution and has homogeneous variance, then perform an independent - samples T - test; if the data follows a normal distribution and has non - homogeneous variance, then perform a modified variance T - test, Welch’s t - test; if the data does not follow a normal distribution, then perform a non - parametric test, Mann - Whitney test. p < 0.05 is considered to have a statistically significant difference. The flowchart of the statistical analysis is as Figure 3 shown. Second, screen out stable and reliable features by evaluating the test - retest reliability level of the image features. Here, the intra - class correlation coefficient ICC(1,1) (given by the following formula) is used to evaluate the test - retest reliability.

[0115]

[0116] where BMS and WMS represent the root mean squares between groups and within groups obtained from Kruskal - Wallis one - way analysis of variance respectively, and k is the number of repeated measurements.

[0117] It should be noted that when analyzing the extracted image features, in addition to performing statistical analysis and repeatability analysis, correlation analysis, comparative analysis, etc. can also be performed on the features.

[0118] Step S400: Obtain an analysis model of the muscle ultrasound image based on the statistically significant features and the reliable features;

[0119] Specifically, statistical and repeatability analysis methods are used to analyze the extracted image features, so as to screen out important features highly related to the disease and stable and reliable features with high test-retest reliability. Based on the statistically different features and the reliable features, an analysis model of the muscle ultrasound image can be obtained.

[0120] Step S500: Input the muscle ultrasound image of the model small animal to be analyzed into the analysis model of the muscle ultrasound image for analysis, and obtain the detection and quantitative evaluation results of the muscle structure of the model small animal to be analyzed.

[0121] Specifically, a muscle ultrasound image analysis system for model small animals is constructed. By inputting the muscle ultrasound image of the model small animal to be analyzed, the corresponding features can be extracted and the analysis results can be obtained, so as to realize the fine detection and quantitative evaluation of the muscle structure of the living model small animal.

[0122] Exemplary device

[0123] This embodiment also provides a muscle ultrasound image analysis device for model small animals. The device includes:

[0124] ROI division module 10, configured to obtain the muscle ultrasound image of the model small animal and perform ROI division on the muscle ultrasound image to obtain the divided muscle ultrasound image;

[0125] Feature extraction module 20, configured to extract muscle morphological features, image frequency features, and image texture features from the divided muscle ultrasound image;

[0126] Feature analysis module 30, configured to perform statistical analysis on the muscle morphological features, image frequency features, and image texture features to obtain statistically different features, and perform repeatability analysis based on test-retest credibility to obtain reliable features;

[0127] Analysis model acquisition module 40, configured to obtain an analysis model of the muscle ultrasound image according to the statistically different features and the reliable features;

[0128] Analysis module 50, configured to input the muscle ultrasound image of the model small animal to be analyzed into the analysis model of the muscle ultrasound image for analysis, and obtain the detection and quantitative evaluation results of the muscle structure of the model small animal to be analyzed.

[0129] In one implementation, the ROI division module 10 includes:

[0130] A pretreatment unit for depilating the hind limbs of a small animal of a test model. In the anesthetized state of the small animal of the test model, the small animal of the test model is placed supine in the center of an experimental board, and its mouth and nose are placed in a pipeline connected to an anesthetic apparatus;

[0131] A detection mode setting unit for setting the detection mode of an ultrasonic imaging system to a musculoskeletal detection mode;

[0132] An ultrasonic probe placement unit for making the long axis of an ultrasonic probe parallel to the Achilles tendon of a small animal of a test model, and keeping the ultrasonic probe placed at a first detection position on the hind limb of the small animal of the test model through setting a mark;

[0133] A muscle ultrasonic image acquisition unit for, in the anesthetized state of the small animal of the test model, applying an ultrasonic gel couplant to ensure acoustic coupling between the ultrasonic probe and the skin, adjusting the ultrasonic probe to optimize the contrast of muscle bundles in an ultrasonic image, and acquiring a muscle ultrasonic image of the small animal of the test model by using a real-time B-mode ultrasonic imaging device based on the detection mode and the first detection position;

[0134] An ROI division unit for graying, cropping the muscle ultrasonic image, and dividing an ROI according to a manually outlined or automatically segmented muscle region to obtain the divided muscle ultrasonic image.

[0135] In one implementation, the feature extraction module 20 includes:

[0136] A Radon transform matrix acquisition unit for performing a normalized Radon transform on the divided muscle ultrasonic image to obtain a Radon transform matrix;

[0137] A Radon transform gradient matrix acquisition unit for taking the gradient of the Radon transform matrix and performing edge enhancement to obtain a Radon transform gradient matrix;

[0138] A first feature acquisition unit for obtaining muscle thickness, muscle fiber length, and pennation angle features according to the Radon transform gradient matrix;

[0139] A muscle morphological feature acquisition unit for obtaining the muscle morphological features according to the muscle thickness, muscle fiber length, and pennation angle features;

[0140] An average frequency analysis feature acquisition unit for extracting the average frequency analysis feature from the divided muscle ultrasonic image; wherein, the calculation formula of the average frequency analysis feature is n, I, and f are respectively the length, power, and frequency of a power density spectrum;

[0141] The first-order statistical feature extraction unit is used to extract first-order statistical features from the segmented muscle ultrasound images based on pixel gray-level distribution calculation; wherein, the first-order statistical features include integrated optical density, mean, standard deviation, variance, skewness, kurtosis, and energy;

[0142] The Haralick feature extraction unit is used to extract Haralick features from the segmented muscle ultrasound images based on gray-level co-occurrence matrix calculation; wherein, the Haralick features include contrast, correlation, energy, entropy, homogeneity, and symmetry;

[0143] The Galloway feature extraction unit is used to extract Galloway features from the segmented muscle ultrasound images based on gray-level run length matrix calculation; wherein, the Galloway features include short run dominance, long run dominance, gray-level non-uniformity, long run non-uniformity, and run percentage;

[0144] The local binary pattern feature extraction unit is used to extract local binary pattern features from the segmented muscle ultrasound images based on the comparison results between the central pixel of the local image region and the neighborhood of the central pixel; wherein, the local binary pattern features include energy and entropy, and the energy is LBP energy =∑ i f i 2 and the entropy is LBP entropy =-∑ i f i 2 log 2 (f i ),where f i represents the corresponding frequency of the i-th block in the local image region;

[0145] The image texture analysis feature extraction unit is used to obtain the image texture analysis features according to the first-order statistical features, the Haralick features, the Galloway features, and the local binary pattern features.

[0146] In one implementation, the feature analysis module 30 includes:

[0147] The first test result acquisition unit is used to perform a normality test on the muscle morphological features, image frequency features, and image texture features to obtain a first test result;

[0148] The second test result acquisition unit is used to perform a homogeneity of variance test on the first test result if the first test result follows a normal distribution to obtain a second test result;

[0149] The first statistically significant feature acquisition unit is used to perform non-parametric tests on the muscle morphological features, image frequency features, and image texture features if the first test result does not follow a normal distribution, so as to obtain the statistically significant features;

[0150] The second statistically significant feature acquisition unit is used to perform independent sample T-tests on the muscle morphological features, image frequency features, and image texture features if the second test result satisfies homoscedasticity, so as to obtain the statistically significant features;

[0151] The third statistically significant feature acquisition unit is used to perform modified variance T-tests on the muscle morphological features, image frequency features, and image texture features if the second test result satisfies heteroscedasticity, so as to obtain the statistically significant features;

[0152] The repeatability test unit is used to perform repeatability tests on the muscle morphological features, image frequency features, and image texture features to obtain a number of test results;

[0153] The reliability acquisition unit is used to evaluate the reliability of the number of test results by using the intraclass correlation coefficient ICC(1,1); where the correlation coefficient is where BMS and WMS are the root mean squares between groups and within groups obtained by Kruskal-Wallis one-way analysis of variance respectively, and k is the number of repeated measurements;

[0154] The reliable feature acquisition unit is used to obtain the reliable features according to the reliability.

[0155] Based on the above embodiments, the present invention also provides an intelligent terminal, and its principle block diagram can be as Figure 5 shown. The intelligent terminal includes a processor, a memory, a network interface, a display screen, and a temperature sensor connected through a system bus. Among them, the processor of the intelligent terminal is used to provide computing and control capabilities. The memory of the intelligent terminal 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 intelligent terminal is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, it realizes a method for extracting and classifying image features based on muscle ultrasound. The display screen of the intelligent terminal can be a liquid crystal display screen or an electronic ink display screen, and the temperature sensor of the intelligent terminal is pre-set inside the intelligent terminal and is used to detect the operating temperature of internal devices.

[0156] Those skilled in the art can understand, Figure 5The principle block diagram shown only shows the block diagram of some structures related to the solution of the present invention, and does not constitute a limitation on the intelligent terminal to which the solution of the present invention is applied. The specific intelligent terminal may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0157] 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 (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0158] In summary, the present invention discloses a method for analyzing muscle ultrasound images of model small animals. The method includes: dividing the ROI of the muscle ultrasound images of the model small animals, and extracting muscle morphological features, image frequency features and image texture features from the divided muscle ultrasound images; performing statistical analysis on the features to obtain statistically different features, and performing repeatability analysis based on test-retest reliability to obtain reliable features; according to the statistically different features and reliable features, obtaining an analysis model of the muscle ultrasound images; inputting the muscle ultrasound images of the model small animals to be analyzed into the analysis model of the muscle ultrasound images for analysis, and obtaining the detection and quantitative evaluation results of the muscle structure of the model small animals to be analyzed. The present invention can collect ultrasound images of the muscle parts of model small animals in vivo, and has the advantages of non-invasive, simple and easy to implement, and can achieve fine detection and visualization of the muscle structure of living model small animals.

[0159] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for analyzing muscle ultrasound images of small animal models, characterized in that, the method includes: Obtaining the muscle ultrasound image of the small animal model, and performing ROI division on the muscle ultrasound image to obtain the divided muscle ultrasound image; Extracting muscle morphological features, image frequency features, and image texture features from the divided muscle ultrasound image; Performing statistical analysis on the muscle morphological features, image frequency features, and image texture features to obtain statistically different features, and performing repeatability analysis based on test-retest reliability to obtain reliable features; According to the statistically different features and the reliable features, an analysis model of the muscle ultrasound image is obtained; Inputting the muscle ultrasound image of the small animal model to be analyzed into the analysis model of the muscle ultrasound image for analysis, and obtaining the detection and quantitative evaluation results of the muscle structure of the small animal model to be analyzed; The performing statistical analysis on the muscle morphological features, image frequency features, and image texture features to obtain statistically different features includes: Performing a normality test on the muscle morphological features, image frequency features, and image texture features to obtain a first test result; If the first test result follows a normal distribution, performing a homogeneity of variance test on the first test result to obtain a second test result; If the first test result does not follow a normal distribution, performing a non-parametric test on the muscle morphological features, image frequency features, and image texture features to obtain the statistically different features; If the second test result satisfies the homogeneity of variance, performing an independent samples T-test on the muscle morphological features, image frequency features, and image texture features to obtain the statistically different features; If the second test result satisfies non-homogeneity of variance, performing a modified variance T-test on the muscle morphological features, image frequency features, and image texture features to obtain the statistically different features.

2. The method for analyzing muscle ultrasound images of small animal models according to claim 1, characterized in that, the obtaining the muscle ultrasound image of the small animal model includes: Performing hair removal on the hind limbs of the test small animal model. Under the anesthetized state of the test small animal model, place the test small animal model supine in the center of the experimental board, and place its mouth and nose in the pipeline connected to the anesthetic apparatus; Setting the detection mode of the ultrasound imaging system to the musculoskeletal detection mode; Aligning the long axis of the ultrasound probe with the Achilles tendon of the test small animal model, and maintaining the placement of the ultrasound probe at the first detection position on the hind limb of the test small animal model by setting a mark; Under the anesthetized state of the test small animal model, applying an ultrasonic gel couplant to ensure acoustic coupling between the ultrasound probe and the skin, adjusting the ultrasound probe to optimize the contrast of muscle bundles in the ultrasound image, and using a real-time B-mode ultrasound imaging device to obtain the muscle ultrasound image of the test small animal model based on the detection mode and the first detection position.

3. The method for analyzing muscle ultrasound images of small animal models according to claim 1, characterized in that, Performing ROI division on the muscle ultrasound image to obtain the divided muscle ultrasound image includes: Grayscale processing, cropping the muscle ultrasound image, and dividing the ROI according to the manually outlined or automatically segmented muscle region to obtain the divided muscle ultrasound image.

4. The method for analyzing muscle ultrasound images of model small animals according to claim 1, characterized in that extracting muscle morphological features and image frequency features from the divided muscle ultrasound image includes: Performing a Radon transform on the divided muscle ultrasound image to obtain a Radon transform matrix; Calculating the gradient of the Radon transform matrix and performing edge enhancement to obtain a Radon transform gradient matrix; Obtaining muscle thickness, muscle fiber length, and pennation angle features based on the Radon transform gradient matrix; Obtaining the muscle morphological features based on the muscle thickness, muscle fiber length, and pennation angle features; Extract the average frequency analysis features from the divided muscle ultrasound images; wherein, the calculation formula of the average frequency analysis features is n, I, and f are the length, power, and frequency of the power density spectrum respectively, and i is an index variable used to represent the number of the frequency component, I i represents the power corresponding to the i-th frequency component, f i represents the i-th frequency component.

5. The method for analyzing muscle ultrasound images of model small animals according to claim 1, characterized in that extracting image texture analysis features from the divided muscle ultrasound image includes: Extracting first-order statistical features from the divided muscle ultrasound image based on pixel gray-level distribution calculation; wherein, the first-order statistical features include integrated optical density, mean, standard deviation, variance, skewness, kurtosis, and energy; Extracting Haralick features from the divided muscle ultrasound image based on gray-level co-occurrence matrix calculation; wherein, the Haralick features include contrast, correlation, energy, entropy, homogeneity, and symmetry; Extracting Galloway features from the divided muscle ultrasound image based on gray-level run length matrix calculation; wherein, the Galloway features include short run dominance, long run dominance, gray-level non-uniformity, long run non-uniformity, and run percentage; Extract local binary pattern features from the segmented muscle ultrasound images based on the comparison results between the central pixel of a local region of the image and the neighborhood of the central pixel; wherein, the local binary pattern features include energy and entropy, and the energy is The entropy is i is an index variable used to represent the pattern number, P represents the number of patterns in the local binary pattern, and f i represents the corresponding frequency of the i-th block in the local region of the image; Obtaining the image texture analysis features based on the first-order statistical features, the Haralick features, the Galloway features, and the local binary pattern features.

6. The method for analyzing muscle ultrasound images of model small animals according to claim 5, characterized in that performing repeatability analysis based on test-retest reliability to obtain reliable features includes: Performing repeatability tests on the muscle morphological features, image frequency features, and image texture features to obtain a number of test results; The reliability of the several test results is evaluated using the intraclass correlation coefficient ICC(1,1); wherein, the correlation coefficient is wherein, BMS and WMS are the root mean squares between groups and within groups obtained by Kruskal-Wallis one-way analysis of variance, respectively, and k is the number of repeated measurements; Obtaining the reliable features based on the reliability.

7. A device for analyzing muscle ultrasound images of model small animals, characterized in that the device includes: An ROI division module for acquiring the muscle ultrasound image of a model small animal and performing ROI division on the muscle ultrasound image to obtain the divided muscle ultrasound image; A feature extraction module for extracting muscle morphological features, image frequency features, and image texture features from the divided muscle ultrasound image; A feature analysis module, configured to perform statistical analysis on the muscle morphological features, image frequency features, and image texture features to obtain statistically different features, and perform repeatability analysis based on test-retest reliability to obtain reliable features; An analysis model acquisition module, configured to obtain an analysis model of the muscle ultrasound image according to the statistically different features and the reliable features; An analysis module, configured to input the muscle ultrasound image of the model small animal to be analyzed into the analysis model of the muscle ultrasound image for analysis, and obtain the detection and quantitative evaluation results of the muscle structure of the model small animal to be analyzed; The feature analysis module includes: A first test result acquisition unit, configured to perform a normality test on the muscle morphological features, image frequency features, and image texture features to obtain a first test result; A second test result acquisition unit, configured to perform a homogeneity of variance test on the first test result if the first test result follows a normal distribution, to obtain a second test result; A first statistically different feature acquisition unit, configured to perform a non-parametric test on the muscle morphological features, image frequency features, and image texture features if the first test result does not follow a normal distribution, to obtain the statistically different features; A second statistically different feature acquisition unit, configured to perform an independent samples T-test on the muscle morphological features, image frequency features, and image texture features if the second test result satisfies the homogeneity of variance, to obtain the statistically different features; A third statistically different feature acquisition unit, configured to perform a modified variance T-test on the muscle morphological features, image frequency features, and image texture features if the second test result satisfies non-homogeneity of variance, to obtain the statistically different features.

8. An intelligent terminal, characterized in that the intelligent terminal includes a memory, a processor, and a program for analyzing the muscle ultrasound image of a model small animal stored in the memory and executable on the processor. When the processor executes the program for analyzing the muscle ultrasound image of a model small animal, the steps of the method for analyzing the muscle ultrasound image of a model small animal according to any one of claims 1-6 are implemented.

9. A computer-readable storage medium, characterized in that a program for analyzing the muscle ultrasound image of a model small animal is stored on the computer-readable storage medium. When the program for analyzing the muscle ultrasound image of a model small animal is executed by a processor, the steps of the method for analyzing the muscle ultrasound image of a model small animal according to any one of claims 1-6 are implemented.

Images