Brain image analysis method and system

Abstract

The present invention discloses a brain image analysis method and system, wherein the method comprises: obtaining a weighted brain magnetic resonance image of a user within a certain time period, extracting first data information of the weighted brain magnetic resonance image, constructing a first data set, wherein the first data information contains different features, extracting time series information associated with the first data information, constructing a second data set with the first data set and the time series information, analyzing the weighted brain magnetic resonance image based on the second data set, presenting the analysis results, and marking abnormal areas. By extracting features from the weighted brain magnetic resonance image and introducing time series information, the features of the image are associated with time, so that the process of feature changes over time can be obtained, and the results obtained by analysis are more accurate than static images, thereby improving the accuracy of feature recognition, and further improving the accuracy of abnormal area marking.

Description

Technical Field

[0001] The present invention belongs to the field of medical image technology, and in particular relates to a brain image analysis method and system. Background Art

[0002] Magnetic resonance imaging is a new medical imaging technology that uses the principle of nuclear magnetic resonance. It has excellent diagnostic capabilities for solid organs such as the brain, thyroid, liver, gallbladder, spleen, kidney, pancreas, adrenal glands, uterus, ovaries, prostate, heart and large blood vessels. Compared with other auxiliary examination methods, nuclear magnetic resonance has the advantages of more imaging parameters, faster scanning speed, higher tissue resolution and clearer images. It can help doctors detect early lesions that are not easy to detect. It is a common detection method for early screening of tumors, heart disease and cerebrovascular diseases.

[0003] Magnetic resonance imaging (MRI) is a multi-parameter imaging method. For the convenience of image analysis, it is hoped that the grayscale of a frame of MRI image is mainly determined by a specific imaging parameter. This type of image, which is intended to highlight a certain tissue characteristic parameter, is called a "weighted image". For example, when it is mainly determined by T1, it is a T1-weighted image; when it is mainly determined by T2, it is a T2-weighted image, and so on.

[0004] Traditionally, when analyzing MRI images, only static MRI images are analyzed. Since MRI image data is dynamic, feature recognition based only on static MRI images is not accurate enough. Summary of the invention

[0005] In view of the defects existing in the above-mentioned prior art, the present invention provides a brain image analysis method, which can solve the problem of low accuracy of image recognition. The method comprises the following steps:

[0006] Step S101, obtaining a weighted brain magnetic resonance image of a user within a certain period of time;

[0007] Step S103, extracting first data information of the weighted brain magnetic resonance image to construct a first data set, wherein the first data information includes different features;

[0008] Step S105: extracting time series information associated with the first data information;

[0009] Step S107, constructing a second data set by combining the first data set and the time series information;

[0010] Step S109, analyzing the weighted brain magnetic resonance image based on the second data set;

[0011] Step S111: present the analysis results and mark the abnormal areas.

[0012] In a certain embodiment, the first data information includes: one or more of: local binary pattern features, grayscale features, oriented gradient histogram features, edge features, linear features, center features and diagonal features.

[0013] In one embodiment, the oriented gradient histogram feature is extracted by the following steps:

[0014] Grayscale the weighted brain magnetic resonance image at a certain moment, and the grayscale image is a three-dimensional image of x, y, z, where x and y represent two-dimensional coordinates and z represents the grayscale value;

[0015] Using a Gamma correction method to standardize the color space of the grayscale image;

[0016] Calculate the gradient of each pixel in the normalized image;

[0017] Dividing the standardized image into a number of minimum connected regions, each of which contains 3×3 pixels;

[0018] Count the gradient histogram of each minimum connected area;

[0019] Every n minimum connected regions are combined into a square region, and the features of all the minimum connected regions in each square region are connected in series to obtain the first feature of the square region;

[0020] The first features of all the square areas in the weighted brain magnetic resonance image are connected in series to obtain the directional gradient histogram feature of the weighted brain magnetic resonance image.

[0021] In one embodiment, the gradient includes magnitude and direction.

[0022] In one embodiment, the local binary pattern feature is extracted by the following steps:

[0023] Dividing the image into a plurality of first regions based on the detection window;

[0024] For each pixel in the first area, the grayscale values ​​of the eight adjacent pixels are compared with the pixel. If the surrounding pixel values ​​are greater than the central pixel value, the position of the pixel is marked as 1, otherwise it is marked as 0.

[0025] calculating a histogram of each first region;

[0026] Performing normalization processing on the histogram;

[0027] The obtained statistical histograms of each first region are connected to form a feature vector, which is a local binary pattern feature vector.

[0028] In one embodiment, the square area is an area consisting of 16×16 pixels.

[0029] In one embodiment, an integral image is used to extract edge features, linear features, center features, and diagonal features.

[0030] In one embodiment, the step S109 includes:

[0031] Assume that n sample data X={x1,...,x n}, where the sample data comes from the second data set, the dimension of the feature is D, and the sample is x i , i=1,2,...,n, jitter each discrete feature of each sample separately, after jitter x←x+random(-0.1,0.1), the jittered data set is recorded as X';

[0032] Randomly select d features in X', with a total of There are 3 combinations. Accordingly, the original data set is divided into m subsets, denoted as S j , where j = 1, 2, ..., m;

[0033] In each subset S j In the example, we calculate each sample x i The presentation value s(i,j);

[0034] Randomly select S j One of the d features q c , c=1,2,...,d, randomly select the dividing line for segmentation, and divide the data set S j Divide into 2 categories and record the number of data in each category C clas，1 ;

[0035] Randomly extract a feature, randomly select the dividing line for segmentation, and divide S j Divided into 4 data sets, record the number of data of each category C clas,2 ;

[0036] Iterate until each data point is classified into a different class, where C clas,num Represents the set of the number of each category after the num-th segmentation.

[0037] In one embodiment, the step S109 further includes:

[0038] Calculate sample x i The presentation value s(i,j);

[0039] Starting from the split, if in C clas,num Medium i The number of classes occupied is 1, then xi If x is 1 type alone, then i In S j The total number of times the samples are segmented is num times, and the number of times all samples are segmented is P = {p1, p2, ..., p i}, where i = 1, 2, ..., n;

[0040] Then x i The presentation value

[0041] Calculate the total presentation value of each sample as s(i),

[0042] in i={1,2,...,n}, j={1,2,...,m};

[0043] Assuming the threshold is α, the set of abnormal presentation values ​​is:

[0044] Set {k}, if s(k)>Quartile{s i ,α}, where i={1,2,...,n}, Quartile{s i ,α} represents the set s i ={s1,s2,...,s n}, then the set {k} is the sequence number of the abnormal samples.

[0045] In a certain embodiment, the abnormal sample serial number indicates a sample that is highly suspected of being abnormal. When multiple abnormal samples are connected, they appear as an abnormal area.

[0046] Compared with the prior art, the present invention obtains the weighted brain magnetic resonance image of the user within a certain period of time, and extracts the first data information of the weighted brain magnetic resonance image, thereby constructing a first data set, the first data information contains different features, and then extracts the time series information associated with the first data information, and constructs a second data set with the first data set and the time series information, and analyzes the weighted brain magnetic resonance image based on the second data set, presents the analysis results, and marks the abnormal area. By extracting features from the weighted brain magnetic resonance image, introducing the time series information, and associating the features of the image with time, the process of feature changes over time can be obtained, and the results obtained by analysis are more accurate than static images, thereby improving the accuracy of feature recognition, and then improving the accuracy of abnormal area marking. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] By reading the detailed description below with reference to the accompanying drawings, the above and other purposes, features and advantages of the exemplary embodiments of the present disclosure will become readily understood. In the accompanying drawings, several embodiments of the present disclosure are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0048] Figure 1 is a flow chart showing a brain image analysis method according to an embodiment of the present invention; and

[0049] Figure 2 is a schematic diagram showing a brain image analysis system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical scheme and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings. 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 creative work are within the scope of protection of the present invention.

[0051] The terms used in the embodiments of the present invention are only for the purpose of describing specific embodiments, and are not intended to limit the present invention. The singular forms "a", "said" and "the" used in the embodiments of the present invention and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings, and "multiple" generally includes at least two.

[0052] The optional embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0053] Embodiment 1

[0054] like Figure 1 As shown, the present invention discloses a brain image analysis method, comprising the following steps:

[0055] Step S101, obtaining a weighted brain magnetic resonance image of a user within a certain period of time;

[0056] Step S103, extracting first data information of the weighted brain magnetic resonance image to construct a first data set, wherein the first data information includes different features;

[0057] Step S105: extracting time series information associated with the first data information;

[0058] Step S107, constructing a second data set by combining the first data set and the time series information;

[0059] Step S109, analyzing the weighted brain magnetic resonance image based on the second data set;

[0060] Step S111: present the analysis results and mark the abnormal areas.

[0061] Compared with the prior art, the present invention obtains the weighted brain magnetic resonance image of the user within a certain period of time, and extracts the first data information of the weighted brain magnetic resonance image, thereby constructing a first data set, the first data information contains different features, and then extracts the time series information associated with the first data information, and constructs a second data set with the first data set and the time series information, and analyzes the weighted brain magnetic resonance image based on the second data set, presents the analysis results, and marks the abnormal area. By extracting features from the weighted brain magnetic resonance image, introducing the time series information, and associating the features of the image with time, the process of feature changes over time can be obtained, and the results obtained by analysis are more accurate than static images, thereby improving the accuracy of feature recognition, and then improving the accuracy of abnormal area marking.

[0062] In one embodiment, the first data information includes: one or more of local binary pattern features, grayscale features, oriented gradient histogram features, edge features, linear features, center features and diagonal features.

[0063] Among them, Local Binary Patterns (LBP) is a visual operator used for classification in the field of computer vision. It is a special case of the Texture Spectrum model proposed in 1990. LBP was first proposed in 1994 and has since been regarded as a powerful feature in texture classification; at the same time, if LBP is combined with the Histogram of Oriented Gradients (HOG) operator, the detection performance can be greatly improved in some data sets. Grayscale features are extracted after the image is grayed. The Histogram of Oriented Gradients feature descriptor can convert a 3-channel color image into a feature vector of a certain length. By applying an edge detector to a given image, a series of continuous curves can be obtained to represent the boundaries of the object. Therefore, the result obtained by applying the edge detection algorithm will greatly reduce the amount of image data, thereby filtering out a lot of information we don’t need, leaving the important structure of the image, and the work to be processed is greatly simplified. Linear features refer to features with a degree of 1. The impact of these features on the results satisfies the addition principle, that is, the whole is equal to the sum of the parts. Edge features, linear features, center features, and diagonal features reflect the grayscale changes of the image. By extracting these features of the weighted brain magnetic resonance image, feature analysis can be performed on the weighted brain magnetic resonance image, thereby improving image recognition accuracy.

[0064] In one embodiment, the oriented gradient histogram feature is extracted by the following steps:

[0065] Grayscale the weighted brain magnetic resonance image at a certain moment, and the grayscale image is a three-dimensional image of x, y, z, where x and y represent two-dimensional coordinates and z represents the grayscale value;

[0066] Using a Gamma correction method to standardize the color space of the grayscale image;

[0067] Calculate the gradient of each pixel in the normalized image;

[0068] Dividing the standardized image into a number of minimum connected regions, each of which contains 3×3 pixels;

[0069] Count the gradient histogram of each minimum connected area;

[0070] Every n minimum connected regions are combined into a square region, and the features of all the minimum connected regions in each square region are connected in series to obtain the first feature of the square region;

[0071] The first features of all the square areas in the weighted brain magnetic resonance image are connected in series to obtain the directional gradient histogram feature of the weighted brain magnetic resonance image.

[0072] Among them, the directional gradient histogram feature descriptor can convert a 3-channel color image into a feature vector of a certain length. By performing a series of processing on the weighted brain magnetic resonance image, the directional gradient histogram feature of the weighted brain magnetic resonance image is obtained. The gradient (x and y derivatives) of the image is very useful because the gradient amplitude around the edges and corners (areas with sudden changes in intensity) is large, and the edges and corners contain more information about the shape of the object than the flat areas.

[0073] In one embodiment, the gradient includes magnitude and direction.

[0074] At each pixel, the gradient has a magnitude and a direction. The x-direction gradient map will strengthen the vertical edge features, and the y-direction gradient map will strengthen the horizontal edge features. This allows useful features (contours) to be retained and irrelevant and unimportant information to be removed.

[0075] In one embodiment, the local binary pattern feature is extracted by the following steps:

[0076] Dividing the image into a plurality of first regions based on the detection window;

[0077] For each pixel in the first area, the grayscale values ​​of the eight adjacent pixels are compared with the pixel. If the surrounding pixel values ​​are greater than the central pixel value, the position of the pixel is marked as 1, otherwise it is marked as 0.

[0078] calculating a histogram of each first region;

[0079] Performing normalization processing on the histogram;

[0080] The obtained statistical histograms of each first region are connected to form a feature vector, which is a local binary pattern feature vector.

[0081] Among them, Local Binary Patterns (LBP) is a visual operator used for classification in the field of computer vision. It is a special case of the Texture Spectrum model proposed in 1990. LBP was first proposed in 1994 and has since been regarded as a powerful feature in texture classification; at the same time, if LBP is combined with the Histogram of Directed Gradients (HOG) operator, the detection performance can be greatly improved in some data sets. The local binary pattern feature vector is obtained by processing the square area divided according to the detection window. Obtaining the local binary pattern feature vector can improve the detection performance, thereby improving the accuracy of image recognition.

[0082] In one embodiment, the square area is an area consisting of 16×16 pixels.

[0083] Among them, dividing the square area into an area consisting of 16×16 pixels can ensure that there are at least 8 pixels next to each pixel, thereby ensuring the generation of an 8-bit binary number, that is, obtaining the LBP value of the center pixel of the window.

[0084] In one embodiment, an integral image is used to extract edge features, linear features, center features, and diagonal features.

[0085] Among them, the integral graph is a fast algorithm that can calculate the sum of pixels in all regions of the image by traversing the image only once, which greatly improves the efficiency of image feature value calculation. The main idea of ​​the integral graph is to save the sum of pixels in the rectangular area formed by the image from the starting point to each point in the memory as an element of an array. When calculating the sum of pixels in a certain area, you can directly index the elements of the array without recalculating the sum of pixels in this area, thereby speeding up the calculation (this has a corresponding name, called dynamic programming algorithm). The integral graph can use the same time (constant time) to calculate different features at multiple scales, thus greatly improving the detection speed. The edge features, linear features, center features and diagonal features reflect the grayscale changes of the image.

[0086] In one embodiment, the step S109 includes:

[0087] Assume that n sample data X={x1,...,x n}, where the sample data comes from the second data set, the dimension of the feature is D, and the sample is xi , i=1,2,...,n, jitter each discrete feature of each sample separately, after jitter x←x+random(-0.1,0.1), the jittered data set is recorded as X';

[0088] Randomly select d features in X', with a total of There are 3 combinations. Accordingly, the original data set is divided into m subsets, denoted as S j , where j = 1, 2, ..., m;

[0089] In each subset S j In the example, we calculate each sample x i The presentation value s(i,j);

[0090] Randomly select S j One of the d features q c , c=1,2,...,d, randomly select the dividing line for segmentation, and divide the data set S j Divide into 2 categories and record the number of data in each category C clas，1 ;

[0091] Randomly extract a feature, randomly select the dividing line for segmentation, and divide S j Divided into 4 data sets, record the number of data of each category C clas,2 ;

[0092] Iterate until each data point is classified into a different class, where C clas,num Represents the set of the number of each category after the num-th segmentation.

[0093] In one embodiment, the step S109 further includes:

[0094] Calculate sample x i The presentation value s(i,j);

[0095] Starting from the split, if in C clas,num Medium i The number of classes occupied is 1, then x i If x is 1 type alone, then i In S j The total number of times the samples are segmented is num times, and the number of times all samples are segmented is P = {p1, p2, ..., p i}, where i = 1, 2, ..., n;

[0096] Then x i The presentation value

[0097] Calculate the total presentation value of each sample as s(i),

[0098] in i={1,2,...,n}, j={1,2,...,m};

[0099] Assuming the threshold is α, the set of abnormal presentation values ​​is:

[0100] Set {k}, if s(k)>Quartile{s i ,α}, where i={1,2,...,n}, Quartile{s i ,α} represents the set s i ={s1,s2,...,s n}, then the set {k} is the sequence number of the abnormal samples.

[0101] In a certain embodiment, the abnormal sample serial number indicates a sample that is highly suspected of being abnormal. When multiple abnormal samples are connected, they appear as an abnormal area.

[0102] Embodiment 2

[0103] like Figure 2 As shown, the present invention also proposes a brain image analysis system, which includes:

[0104] An acquisition module 202 is used to acquire a weighted brain magnetic resonance image of a user within a certain period of time;

[0105] A first extraction module 204 is used to extract first data information of the weighted brain magnetic resonance image to construct a first data set, wherein the first data information includes different features;

[0106] A second extraction module 206, configured to extract time series information associated with the first data information;

[0107] A construction module 208 is used to construct a second data set by combining the first data set and the time series information;

[0108] an analysis module 210, configured to analyze the weighted brain magnetic resonance image based on the second data set;

[0109] The presentation module 212 is used to present the analysis results and mark abnormal areas.

[0110] Compared with the prior art, the present invention obtains the weighted brain magnetic resonance image of the user within a certain period of time, and extracts the first data information of the weighted brain magnetic resonance image, thereby constructing a first data set, the first data information contains different features, and then extracts the time series information associated with the first data information, and constructs a second data set with the first data set and the time series information, and analyzes the weighted brain magnetic resonance image based on the second data set, presents the analysis results, and marks the abnormal area. By extracting features from the weighted brain magnetic resonance image and introducing the time series information, the image features are associated with time, and the analysis results are more accurate than static images, thereby improving the accuracy of image recognition.

[0111] Embodiment 3

[0112] An embodiment of the present disclosure provides a non-volatile computer storage medium, wherein the computer storage medium stores computer executable instructions, and the computer executable instructions can execute the method steps described in the above embodiment.

[0113] It should be noted that the computer-readable medium disclosed above may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which a computer-readable program code is carried. This propagated data signal may take a variety of forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0114] The computer-readable medium may be included in the electronic device, or may exist independently without being incorporated into the electronic device.

[0115] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (AN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0116] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0117] The units involved in the embodiments described in the present disclosure may be implemented by software or hardware, wherein the name of a unit does not, in some cases, limit the unit itself.

[0118] The above introduces the preferred embodiments of the present invention, which is intended to make the spirit of the present invention clearer and easier to understand, but is not intended to limit the present invention. All modifications, substitutions, and improvements made within the spirit and principles of the present invention should be included in the scope of protection outlined by the claims attached to the present invention.

Claims

1. A brain image analysis method, characterized in that: The following steps are involved: Step S101, obtaining a weighted brain magnetic resonance image of a user within a certain period of time; Step S103, extracting first data information of the weighted brain magnetic resonance image to construct a first data set, wherein the first data information includes different features; Step S105: extracting time series information associated with the first data information; Step S107, constructing a second data set by combining the first data set and the time series information; Step S109, analyzing the weighted brain magnetic resonance image based on the second data set; Step S111, presenting the analysis results and marking the abnormal areas; Wherein the step S109 comprises: assumed Sample data , where the sample data comes from the second data set and the dimension of the feature is , and the sample is , , jitter each discrete feature of each sample separately, after jitter , the jittered data set is recorded as ; exist Choose any of Features, total combinations, accordingly, the original data set is divided into subset, denoted as ,in ; In each subset , calculate each sample The presentation value ; Random draw of One feature among features , , randomly select the dividing line to split the data set Divide into 2 categories and record the number of data in each category ; Randomly extract a feature and randomly select a dividing line for segmentation. Divided into 4 data sets, recording the number of data in each category ; Iterate until each data point is classified into a different class, where Indicates The set of the number of each type after the split; The step S109 further includes: Calculation Sample The presentation value ; Starting from the split, if middle If the number of classes occupied is 1, then If it is a single category 1, exist The total number of times in is divided is times, the number of splits of all samples is recorded as ,in ; but The presentation value ; Calculate the total presentation value of each sample as , ,in , , ; Set the threshold to , then the set of abnormal presentation values ​​is: gather ,if ,in , Representing a collection of The set {k} is the sequence number of the abnormal samples.

2. The brain image analysis method according to claim 1, characterized in that: The first data information includes: one or more of local binary pattern features, grayscale features, directional gradient histogram features, edge features, linear features, center features and diagonal features.

3. The brain image analysis method according to claim 2, characterized in that: The oriented gradient histogram feature is extracted using the following steps: Grayscale the weighted brain magnetic resonance image at a certain moment, and the grayscale image is a three-dimensional image of x, y, z, where x and y represent two-dimensional coordinates and z represents the grayscale value; Using a Gamma correction method to standardize the color space of the grayscale image; Calculate the gradient of each pixel in the normalized image; Dividing the standardized image into a number of minimum connected regions, each of which contains 3×3 pixels; Count the gradient histogram of each minimum connected area; Every n minimum connected regions are combined into a square region, and the features of all the minimum connected regions in each square region are connected in series to obtain the first feature of the square region; The first features of all the square areas in the weighted brain magnetic resonance image are connected in series to obtain the directional gradient histogram feature of the weighted brain magnetic resonance image.

4. The brain image analysis method according to claim 3, characterized in that: The gradient includes magnitude and direction.

5. The brain image analysis method according to claim 2, characterized in that: The local binary pattern features are extracted using the following steps: Dividing the image into a plurality of first regions based on the detection window; For a pixel in each first area, the grayscale values ​​of the eight adjacent pixels are compared with it. If the surrounding pixel values ​​are greater than the central pixel value, the position of the pixel is marked as 1, otherwise it is marked as 0; calculating a histogram of each first region; Performing normalization processing on the histogram; The obtained statistical histograms of each first region are connected to form a feature vector, which is a local binary pattern feature vector.

6. The brain image analysis method according to claim 5, characterized in that: The square area is an area consisting of 16×16 pixels.

7. The brain image analysis method according to claim 2, characterized in that: Integral images are used to extract edge features, linear features, center features and diagonal features.

8. A brain image analysis system, characterized in that: For applying the brain image analysis method according to any one of claims 1 to 7, the system comprises: An acquisition module, used for acquiring a weighted brain magnetic resonance image of a user within a certain period of time; A first extraction module, configured to extract first data information from the weighted brain magnetic resonance image to construct a first data set, wherein the first data information includes different features; A second extraction module, used to extract time series information associated with the first data information; A construction module, used to construct a second data set by combining the first data set and the time series information; an analysis module, configured to analyze the weighted brain magnetic resonance image based on the second data set; The presentation module is used to present the analysis results and mark abnormal areas.

Images