Brain disease diagnosis and brain disease information provision system

Abstract

The brain disease diagnosis and brain disease information provision system according to the present invention is characterized by comprising: an image input unit that receives a brain image; a brain disease diagnosis unit that diagnoses a brain disease and generates a diagnosis file through a brain image provided from the image input unit; a brain neural pathway tracking unit that tracks a brain neural pathway and generates a tracking file through a brain image provided from the image input unit; and an information provision unit that generates a brain disease information file through the diagnosis file of the brain disease diagnosis unit and the tracking file of the brain neural pathway tracking unit, and provides visual and linguistic information about the brain disease through the brain disease information file.

Description

Technology Field

[0001] The present invention relates to a brain disease diagnosis and brain disease information provision system, and more specifically, to a brain disease diagnosis and brain disease information provision system for diagnosing a brain disease through input brain images and generating and providing visual and linguistic information about the brain disease based on the diagnosis results. Background Technology

[0003] As we enter a super-aged society, the prevalence of brain diseases is rapidly increasing, and accordingly, the need for rapid and accurate diagnosis is being emphasized.

[0004] In particular, rapid and precise diagnosis and treatment are essential for major brain diseases such as brain tumors, strokes, and Alzheimer's disease.

[0005] However, existing brain imaging-based diagnostic methods were time-consuming and suffered from inconsistencies in results depending on the proficiency of the medical staff.

[0006] In addition, according to Korean Registered Patent No. 10-1797501, an analysis device is provided for diagnosing brain disease by comparing brain images of an experimental group suspected of having brain disease with brain images of a normal control group.

[0007] However, this device had a problem in that it only provided diagnostic results for brain diseases and could not provide the type of disease or additional information. Prior art literature

[0009] Korean Registered Patent No. 10-1797501 (Published Nov. 16, 2017) The problem to be solved

[0010] The objective of the present invention is to provide a brain disease diagnosis and brain disease information provision system for rapidly and accurately diagnosing major brain diseases, including brain tumors, strokes, and Alzheimer's disease, in order to solve the problems described above.

[0011] Another objective of the present invention is to provide a brain disease diagnosis and brain disease information provision system for providing the type of diagnosed brain disease and additional information thereon. means of solving the problem

[0013] To achieve the above objective, the brain disease diagnosis and brain disease information provision system according to the present invention is characterized by comprising: an image input unit that receives a brain image; a brain disease diagnosis unit that diagnoses a brain disease and generates a diagnosis file through a brain image provided from the image input unit; a brain neural pathway tracking unit that tracks a brain neural pathway and generates a tracking file through a brain image provided from the image input unit; and an information provision unit that generates a brain disease information file through the diagnosis file of the brain disease diagnosis unit and the tracking file of the brain neural pathway tracking unit, and provides visual and linguistic information about the brain disease through the brain disease information file.

[0014] In addition, the brain imaging is characterized by being magnetic resonance imaging.

[0015] In addition, the brain disease diagnosis unit is characterized by diagnosing brain diseases through T1-weighted images of magnetic resonance imaging and generating a diagnosis file.

[0016] In addition, the brain neural pathway tracking unit is characterized by tracking brain neural pathways through diffusion tensor imaging among magnetic resonance imaging and generating a tracking file.

[0017] In addition, the brain disease information file is characterized by being composed of a 3D modeling file capable of 3D printing and a prompt file that is input into a generative language model.

[0018] In addition, the information provider is characterized by providing visual information about brain diseases through a 3D modeling file and providing linguistic information about brain diseases by inputting a prompt file into a generative language model. Effects of the invention

[0020] As described above, the brain disease diagnosis and brain disease information provision system according to the present invention has the effect of enabling early diagnosis and treatment of brain diseases by diagnosing a brain disease through an input brain image and generating and providing visual and linguistic information about the brain disease based on the diagnosis results.

[0021] In addition, according to the brain disease diagnosis and brain disease information provision system of the present invention, by diagnosing a brain disease through an input brain image and generating and providing visual and linguistic information about the brain disease based on the diagnosis results, there is an effect that enables medical staff and patients to better understand the diagnosis results. Brief explanation of the drawing

[0023] FIG. 1 is a simplified diagram illustrating the configuration of a brain disease diagnosis and brain disease information provision system according to the present invention. FIG. 2 is an exemplary diagram briefly illustrating a brain image input to a brain disease diagnosis and brain disease information provision system according to the present invention. FIG. 3 is a diagram illustrating the diagnostic process of the brain disease diagnosis unit in the brain disease diagnosis and brain disease information provision system according to the present invention. FIG. 4 is a diagram illustrating the tracking process of the brain nerve pathway tracking unit in the brain disease diagnosis and brain disease information provision system according to the present invention. FIG. 5 is a diagram illustrating the information provision process of the information provision unit in a brain disease diagnosis and brain disease information provision system according to the present invention. FIG. 6 is a diagram briefly illustrating a graphical user interface of a brain disease diagnosis and brain disease information provision system according to the present invention. FIG. 7 is an exemplary diagram illustrating a 3D printing result of a brain disease diagnosis and brain disease information provision system according to the present invention. Specific details for implementing the invention

[0024] Specific structural or functional descriptions of embodiments according to the concept of the present invention disclosed herein are provided merely for the purpose of explaining embodiments according to the concept of the present invention, and embodiments according to the concept of the present invention may be implemented in various forms and are not limited to the embodiments described herein.

[0025] Embodiments according to the concept of the present invention may be subject to various modifications and may take various forms; therefore, embodiments are illustrated in the drawings and described in detail in this specification. However, this is not intended to limit the embodiments according to the concept of the present invention to specific disclosed forms, and includes all modifications, equivalents, or substitutions that fall within the spirit and scope of the present invention.

[0027] Hereinafter, preferred embodiments of the present invention will be described with reference to the attached drawings.

[0029] FIG. 1 is a simplified diagram illustrating the configuration of a brain disease diagnosis and brain disease information provision system according to the present invention; FIG. 2 is a simplified example diagram illustrating a brain image input to a brain disease diagnosis and brain disease information provision system according to the present invention; FIG. 3 is a diagram illustrating the diagnosis process of a brain disease diagnosis unit in a brain disease diagnosis and brain disease information provision system according to the present invention; FIG. 4 is a diagram illustrating the tracking process of a brain neural pathway tracking unit in a brain disease diagnosis and brain disease information provision system according to the present invention; FIG. 5 is a diagram illustrating the information provision process of an information provision unit in a brain disease diagnosis and brain disease information provision system according to the present invention; FIG. 6 is a simplified diagram illustrating a graphical user interface of a brain disease diagnosis and brain disease information provision system according to the present invention; and FIG. 7 is a simplified example diagram illustrating a 3D printing result of a brain disease diagnosis and brain disease information provision system according to the present invention.

[0031] As illustrated in FIG. 1, the brain disease diagnosis and brain disease information provision system according to the present invention comprises an image input unit (100) that receives a brain image, a brain disease diagnosis unit (200) that diagnoses a brain disease through the input brain image and generates a diagnosis file (10), a brain nerve pathway tracking unit (300) that tracks a brain nerve pathway through the brain image and generates a tracking file (20), and an information provision unit (400) that generates a brain disease information file through the generated diagnosis file (10) and tracking file (20) and provides visual and linguistic information about the brain disease using the brain disease information file.

[0033] The above image input unit (100) receives a brain image captured by magnetic resonance imaging.

[0035] At this time, referring to Fig. 2, the input brain image consists of a T1-weighted image, which can visualize structural details of the tissue based on the time it takes for the magnetization to return to its original state after receiving external magnetic field stimulation, and a Diffusion Tensor Imaging (DTI), which can visualize the structural connectivity and directionality of nerve fibers by measuring the diffusion movement of water molecules within the tissue.

[0037] Here, as illustrated in FIG. 2b, the Diffusion Tensor Imaging is configured to include a bvec file that provides a gradient vector corresponding to the main direction in which each water molecule diffuses, and a bval file that provides the diffusion intensity (b-value) for the direction in which the water molecule diffuses.

[0039] The above brain disease diagnosis unit (200) diagnoses brain disease through a T1-weighted image among brain images.

[0041] At this time, the brain disease diagnosis unit (200) performs a preprocessing step including skull stripping and alignment to the MNI (Montreal Neurological Institute) standard coordinate system on the T1-weighted image, thereby enabling comparison of different brain data in the same coordinate system or group analysis.

[0043] Subsequently, as shown in FIG. 3, the brain disease diagnosis unit (200) has three models for accurate diagnosis of brain disease.

[0045] The first model determines whether there is a lesion in the brain through the first classification step (S1) and classifies brain images based on the presence or absence of the lesion.

[0046] If it is determined that there is a lesion, the second model classifies whether the lesion is a brain tumor or a stroke through the second classification step (S2).

[0047] If it is determined that there are no lesions, the third model classifies whether the brain has Alzheimer's disease or is normal through the third classification step (S3).

[0049] Additionally, if the brain disease diagnosis unit (200) determines that there is a lesion based on the results of the classification steps (S1, S2, S3) described above, it divides the lesion in the corresponding brain through the first division step (A1) and determines its size.

[0050] At this stage, the matrix file generated during alignment with the MNI standard is inversely transformed to identify lesions invading the gray matter region of the brain. The Automated Anatomical Labeling (AAL) atlas is utilized to identify lesions invading the gray matter region, and the Johns Hopkins University (JHU) atlas is used to identify lesions invading the white matter region, thereby determining the location of the lesions and the level of damage by region.

[0052] In addition, if the brain disease diagnosis unit (200) determines that there are no lesions based on the results of the classification steps (S1, S2, S3) described above, it divides 78 brain regions through the second division step (A2), and the six major regions (hippocampus, amygdala, caudate, entorhinal cortex, fusiform gyrus, parahippocampal gyrus) that clearly exhibit the pathological characteristics of Alzheimer's disease are essentially saved as separate files for 3D modeling.

[0054] That is, the brain disease diagnosis unit (200) diagnoses a brain disease from a brain image through all classification steps (S1, S2, S3) and produces a diagnosis file (10) including a file for 3D modeling, damage information by lesion location and region, and brain disease classification information.

[0056] The above-mentioned brain neural pathway tracking unit (300) tracks brain neural pathways through diffusion tensor imaging during brain imaging.

[0058] At this time, the brain neural pathway tracing unit (300) performs a preprocessing step (B1) including skull removal and motion correction on a diffusion tensor imaging including bvec and bval files to prepare a brain image in a state suitable for analysis.

[0060] Subsequently, as illustrated in FIG. 4, the brain neural pathway tracking unit (300) visualizes the brain structure by generating a Fractional Anisotropy (FA) map and a Fractional Anisotropy color map through a map generation step (M1).

[0062] In addition, the above-mentioned brain nerve pathway tracking unit (300) tracks nerve pathways within the entire brain region through the brain nerve pathway tracking step (C1) and extracts the overall appearance of nerve pathways within the brain white matter.

[0064] Next, the brain nerve pathway tracking unit (300) divides 45 major nerve pathways, including the affected side and the unaffected side, using labels from the JHU (Johns Hopkins University) atlas that are aligned with the B0 native space, which is the eigenspace of Diffusion Tensor Imaging, through the brain nerve pathway segmentation step (D1).

[0066] At this time, the brain nerve pathway tracking unit (300) stores information values ​​for the average length, curvature, volume, FA, MD (Mean Diffusivity), AD (Axial Diffusivity), and RD (Radial Diffusivity) of the nerve pathways corresponding to 45 divided data.

[0067] Here, the FA information value of the neural pathway is utilized for neural pathway damage level information.

[0069] That is, the above-mentioned brain nerve pathway tracking unit (300) tracks the brain nerve pathway, identifies information about the brain nerve pathway, and generates a tracking file (20) containing the information.

[0071] Here, a 3D Inception U-Net model is applied that integrates an Inception module, which applies filters of various sizes in parallel to the segmentation model used in the brain disease diagnosis unit (200) and the brain neural pathway tracking unit (300), in order to effectively extract features and implement precise lesion segmentation.

[0073] As illustrated in FIG. 5, the information providing unit (400) generates a brain disease information file including a 3D modeling file (30) that is 3D printable and a prompt file (40) that is input into a generative language model (410) through the diagnostic file (10) and trace file (20) described above.

[0075] At this time, the information providing unit (400) performs the following process to generate a 3D modeling file (30).

[0076] First, based on a diagnostic file (10) containing region and lesion segmentation information, the information providing unit (400) adjusts the resolution of the brain image using voxel information of the T1-weighted image, performs meshing and smoothing, and then upsamples by 2 times.

[0078] In addition, since the 3D modeling result generated as above has a rough surface, the information providing unit (400) converts it into a smooth surface using Laplacian smoothing.

[0080] Secondly, based on a tracking file (20) containing information about the nerve pathway, the information providing unit (400) models each tracked fiber path into a cylinder shape and creates each cylinder in a shape connecting the start point and the end point of the fiber path.

[0082] Subsequently, the information providing unit (400) adjusts the previously set coordinates of each cylinder using the direction vector and length calculated from the Diffusion Tensor Imaging, and thereby generates the total coordinates corresponding to each cylinder and connects them to each other.

[0084] Finally, the information providing unit (400) converts a visually verifiable 3D modeling file (30) into an STL file so that it can be 3D printed.

[0086] In addition, the information providing unit (400) performs the following process to generate a prompt file (40).

[0088] The above information providing unit (400) calculates the degree of nerve pathway damage based on the FA values ​​of the affected side and the unaffected side calculated by the brain nerve pathway tracking unit (300), and stores the brain disease information classified by the brain disease diagnosis unit (200) and information regarding the degree of nerve pathway damage in JSON format.

[0090] At this time, the information providing unit (400) generates a prompt file (40) that is input into a generative language model (410) using Llama 3.1-8B based on information about brain disease and neural pathway damage level information stored in JSON format.

[0092] That is, the information providing unit (400) provides visual information (50) about brain disease through a 3D modeling file (30) and provides verbal information (60) about brain disease through a prompt file (40).

[0094] The user can check visual information (50) and verbal information (60) about brain diseases produced by the information provider (400) through a graphic user interface.

[0096] At this time, in order for the user to check visual information (50) and verbal information (60) about brain disease through a graphic user interface, the following process is performed.

[0098] First, as shown in Fig. 6, the user inputs not only brain images through a graphic user interface but also the user's age and gender.

[0099] Here, the user's age and gender are optional information that does not require entry.

[0101] At this point, once the user inputs all the necessary information, the system begins analyzing brain images based on that information and performs both the brain disease classification and neural pathway tracing described above.

[0103] Secondly, after the brain disease classification and brain nerve pathway tracking are completed, the user can check both visual information (50) and verbal information (60) regarding the brain disease.

[0105] Thirdly, as illustrated in FIG. 7, the user can perform 3D printing operations using the converted 3D modeling file (30).

[0107] Here, the user can print the entire brain, print only the lesion area, or print only brain nerve fiber bundles.

[0109] Finally, users can also engage in question-and-answer sessions regarding the brain's information through a graphical user interface.

[0111] In one example, if a user asks, "What functions do the right middle frontal gyrus and the superior longitudinal fasciculus perform?", the following answer may be obtained.

[0113] The right middle frontal gyrus and the superior terminal bundle are involved in language processing, particularly speech and comprehension, and play a crucial role in executive functions such as decision-making and problem-solving. The right middle frontal gyrus is associated with attention and working memory, while the superior terminal bundle is a white matter pathway connecting the frontal and parietal lobes that is involved in the coordination of movements and the planning of complex motor tasks. Damage to these areas due to a stroke can lead to impairments in language, executive functions, and motor control.

[0115] This project (result) is the result of the Local Government-University Cooperation-based Regional Innovation Project, conducted in 2024 with funding from the Ministry of Education and support from the National Research Foundation of Korea (2022RIS-006).

[0117] As described above, although the present invention has been described with reference to preferred embodiments with reference to the accompanying drawings, it is evident to those skilled in the art that many obvious variations are possible from this description without departing from the scope of the invention. Accordingly, the scope of the invention should be interpreted by the claims described to include examples of such many variations. Explanation of the symbols

[0119] 10 : Diagnostic file 20 : Trace file 30 : 3D modeling file 40 : Prompt file 50: Visual Information 60: Linguistic Information 100 : Video input section 200 : Brain Disease Diagnosis Department 300: Cranial pathway tracing unit 400 : Information Provision Department 410: Generative Language Model

Claims

Claim 1 A brain disease diagnosis and brain disease information provision system characterized by comprising: an image input unit for receiving a brain image; a brain disease diagnosis unit that diagnoses a brain disease through the brain image provided from the image input unit and generates a diagnosis file; a brain nerve pathway tracking unit that tracks a brain nerve pathway through the brain image provided from the image input unit and generates a tracking file; and an information provision unit that generates a brain disease information file through the diagnosis file of the brain disease diagnosis unit and the tracking file of the brain nerve pathway tracking unit, and provides visual and linguistic information about the brain disease through the brain disease information file. Claim 2 A brain disease diagnosis and brain disease information provision system according to claim 1, characterized in that the brain image is a magnetic resonance image. Claim 3 In claim 2, the brain disease diagnosis unit is characterized by diagnosing a brain disease through T1-weighted imaging among magnetic resonance imaging and generating a diagnosis file, in a brain disease diagnosis and brain disease information provision system. Claim 4 In claim 2, the brain neural pathway tracking unit is characterized by tracking brain neural pathways through diffusion tensor imaging among magnetic resonance imaging and generating a tracking file, in a brain disease diagnosis and brain disease information provision system. Claim 5 A brain disease diagnosis and brain disease information provision system according to claim 1, characterized in that the brain disease information file comprises a 3D modeling file capable of 3D printing; and a prompt file input into a generative language model. Claim 6 A brain disease diagnosis and brain disease information provision system according to claim 5, wherein the information providing unit provides visual information about the brain disease through the 3D modeling file and provides linguistic information about the brain disease by inputting the prompt file into the generative language model.

Images