A method for cross-species comparison of brain regions based on gray matter brain regions and fiber tracts

Abstract

The application discloses a brain region homology cross-species comparison method based on gray matter brain regions and fiber bundles, which comprises the following steps: acquiring cross-species brain image data and performing pretreatment; extracting a region of interest from a standard template and performing brain region division to obtain a region of interest division result of different species; using whole brain-based probabilistic fiber tracking to construct connection fingerprints of each subregion of the region of interest and the gray matter brain region; using voxel-based probabilistic fiber tracking to construct connection fingerprints of each subregion of the region of interest and the fiber bundle; obtaining effective gray matter brain region connection fingerprints and fiber bundle connection fingerprints within a species at a group level; calculating similarity or difference between each subregion of the region of interest of different species based on the connection fingerprints to obtain cross-species homology scores; and verifying the method based on the consistency of the homologous subregion to the whole brain functional connection mode. From the perspective of brain mapping and evolution, the method provides a new idea and a new method for cross-species homology comparison of specific brain region subregions by taking the gray matter brain region and the subcortical fiber bundle as feature inputs.

Description

Technical Field

[0001] This invention relates to the field of neuroimaging technology, and to a method for comparing the subregional homology of a region of interest using cross-species cortical brain regions, subcortical nuclei, and fiber tracts. Specifically, it is a cross-species method for comparing brain region homology based on gray matter brain regions and fiber tracts. Background Technology

[0002] Brain diseases are conditions affecting the brain, spinal cord, or nervous system, including stroke, Parkinson's disease, Alzheimer's disease, and meningitis. These diseases can lead to cognitive decline, motor impairments, and emotional problems, and in severe cases, can even be life-threatening. Therefore, the prevention and treatment of brain diseases are crucial for protecting people's health and improving their quality of life. Early detection and intervention of brain diseases, along with appropriate treatment measures, can effectively alleviate patients' suffering, slow disease progression, and improve their quality of life.

[0003] Research has found a close link between many brain diseases and specific subregions of the brain. Different regions of the brain have different functional characteristics, and these regions can be further divided into multiple subregions, each of which may play a specific role in cognitive, emotional, or motor functions.

[0004] Animal models of brain diseases play an irreplaceable role in basic and preclinical research. They help us understand disease mechanisms, including pathophysiological changes, neuronal damage and repair processes; simulate the characteristics of human diseases to understand their occurrence and development; evaluate treatment effects; and test the safety and efficacy of new drugs and treatments. For ethical reasons, conducting experiments directly on humans is unethical or impractical in certain situations. Using animal models allows necessary scientific research to be conducted while adhering to ethical standards.

[0005] Cross-species brain research is crucial for a deeper understanding of the mechanisms underlying brain structure, function, and behavior, and for advancing the field of neuroscience. Studies have shown that changes in cortical volume and functional connectivity between brain regions during primate evolution may provide the structural basis for human-specific functions such as language and tool use. Structural-functional analysis of primates is key to understanding the development of human cognition and behavior. Chimpanzees and macaques, as close relatives on the human evolutionary tree, serve as important references for cross-species brain research, and are significant for exploring the homology of corresponding brain regions and subregions with humans.

[0006] Typically, cross-species comparative studies of brain region homology require prior reference to homologous brain regions between species. Numerous studies have already published widely accepted homologous brain regions across different species. However, rigorous standards are still lacking for subdividing specific brain regions and comparing the homology of their subregions across species. Furthermore, while some studies suggest that white matter fiber tracts can serve as a standard for homology comparison, few studies compare the structural connectivity between subregions and gray matter regions, and their structural connectivity with subcortical fiber tracts, as integral characteristics. Summary of the Invention

[0007] To address the aforementioned issues, the technical solution of this invention is as follows: This invention aims to propose a cross-species comparison method for brain region homology based on gray matter brain regions and fiber tracts, including acquiring standard brain magnetic resonance imaging data from different species and preprocessing them, extracting and dividing brain regions of interest, calculating the connectivity fingerprint between each subregion of the region of interest and the gray matter brain region, calculating the connectivity fingerprint between each subregion of the region of interest and the fiber tract, cross-species homology comparison, and homology verification.

[0008] The standard brain magnetic resonance imaging (MRI) data of different species was acquired and preprocessed. The structural T1w images, functional fMRI images, and diffusion-weighted imaging (DTI) images of the standard brain MRI data of different species were preprocessed to obtain preprocessed structural image data, functional image data, and diffusion-weighted image data.

[0009] The brain regions of interest are extracted and divided. The corresponding regions of interest (ROIs) are extracted from standard brain templates of T1w structural images of different species. Whole-brain probabilistic fiber tracing is applied to the ROIs to obtain the whole-brain structural connectivity of the ROIs, and the ROIs are divided based on clustering methods to obtain the brain region division results of the ROIs of different species.

[0010] The process involves calculating the connectivity fingerprints between each subregion of the region of interest and the gray matter brain region, constructing homology maps of the gray matter brain region in different species, applying probabilistic fiber tracing to obtain the structural connections from each subregion of the region of interest to the whole brain in different species, screening effective connections within species at the group level, and extracting the connectivity fingerprints between each subregion of the region of interest and the gray matter brain region, which includes cortical brain regions and subcortical nuclei.

[0011] The process involves calculating the connection fingerprint between each subregion of the region of interest and the fiber bundle, extracting fiber bundles common to different species as templates, obtaining voxel-based fiber tracking maps of each subregion of the region of interest based on probabilistic fiber tracking, calculating the connection probability value between each subregion of the region of interest and each fiber bundle, filtering effective connections within species at the group level, and extracting the connection fingerprint between each subregion of the region of interest and the fiber bundle.

[0012] The cross-species homology comparison uses the connectivity fingerprints between each subregion of the region of interest and the gray matter brain region and fiber tracts as feature inputs to calculate the similarity or difference between the subregions of different species and score them. The score is used as the homology comparison criterion.

[0013] The cross-species homology verification is based on fMRI functional imaging data to calculate the functional connectivity from each subregion of different species to the whole brain. The whole-brain functional connectivity patterns of homologous subregions of different species are compared, and whether the functional connectivity patterns between species are consistent is used to verify the reliability of this method.

[0014] Furthermore, the acquisition and preprocessing of standard brain magnetic resonance imaging (MRI) data from different species involves using standard brain MRI data from different species, including structural T1w images, functional fMRI images, and diffusion-weighted tensor (DTI) images. The data preprocessing process specifically includes denoising, registration, skull removal, and brain tissue segmentation of the T1w structural images; motion correction, spatial standardization, smoothing, signal extraction, and normalization of the fMRI functional images; and denoising, registration, correction, brain tissue extraction, and diffusion tensor fitting of the diffusion MRI DTI images.

[0015] Furthermore, the extraction of corresponding regions of interest and the segmentation of brain regions mainly involve extracting corresponding regions of interest (ROIs) from standard brain templates of T1w structural images of different species and segmenting these regions. The process mainly includes:

[0016] A) Identify the region of interest (ROI) to be studied and extract the corresponding region from the standard brain template of T1w structural images of different species;

[0017] B) Based on the preprocessed structural image data and diffusion-weighted image data, the extracted region of interest is registered and aligned with the diffusion space. The region of interest in the diffusion space is used as a template mask for seed points. For each voxel in the mask, probabilistic fiber tracing is used to estimate the connection probability, and the connection matrix M between each voxel and the whole brain voxels is obtained.

[0018] C) Downsample the whole-brain connectivity matrix M at each voxel in the region of interest (ROI) to form the voxel connectivity matrix A within the ROI. m×n The number of voxels in the ROI is m, and the number of voxels in the whole brain is n;

[0019] D) Calculate the cross-correlation matrix between the whole-brain connectivity matrices of all voxels within the region of interest (ROI). m×m And used for automatic brain region segmentation;

[0020] E) Apply clustering to the cross-correlation matrix, grouping voxels with higher similarity together to form 2-10 groups and calculating the clustering index (profile coefficient SC, sum of squared errors SSE) for different groups. Select the optimal solution k of the clustering index as the division of the region of interest in different species, and obtain the k subregion division results of the region of interest in different species respectively.

[0021] Furthermore, the main process for calculating the connectivity fingerprint between each subregion of the region of interest and the gray matter brain region includes:

[0022] A) Construct homology maps of gray matter brain regions in different species based on existing T1w structural images and standard brain atlases. Gray matter brain regions include cortical brain regions and subcortical nuclei.

[0023] B) Based on preprocessed structural image data and diffusion-weighted image data, a whole-brain-based probabilistic fiber tracing algorithm was applied to each subregion of different species. The fiber tracing results were then standardized, denoised, and false positives were removed to obtain the structural connectivity from different species subregions to the whole brain. k×n×s The number of subregions is k, the number of whole-brain voxels is n, and the number of individuals is s;

[0024] C) Connecting different species subregions to the whole brain B k×n×s Using a homology map as a template, and taking voxels within each gray matter region of the homology map as a whole, the average structural connectivity from an individual's subregion to the gray matter region is taken as the connectivity strength from the subregion of the region of interest (ROI) to that region. This yields the connectivity fingerprint B from the subregion of the ROI to the cortical region and subcortical nuclei at the individual level. k×j×s The number of subregions is k, the number of gray matter brain regions is j, and the number of individuals is s;

[0025] D) Connectivity fingerprints of the individual-level region of interest (ROI) subregions to cortical brain regions and subcortical nuclei. k×j×s Group-level averaging was performed to obtain the connectivity fingerprint between group-based region of interest (ROI) subregions and gray matter brain regions. k×j The number of subregions is k, and the number of gray matter brain regions is j;

[0026] E) Connectivity fingerprints of cortical regions and subcortical nuclei at the individual level B k×j×s Group-level screening was performed to obtain effective intraspecific connections as features. Based on the screening results, connection fingerprints C' between group-based region of interest (ROI) subregions and gray matter brain regions were extracted. k×j The number of subregions is k, and the number of gray matter brain regions is j;

[0027] Furthermore, the main process for calculating the connection fingerprint between each subregion of the region of interest and the fiber bundle includes:

[0028] A) Based on the common fiber bundles of different species as the basis for cross-species comparison, each fiber bundle is reconstructed for different species. After constructing a mask for each fiber bundle in the MNI space, it is registered to the individual to obtain the mask of the individual fiber bundle.

[0029] B) Using probabilistic fiber tracing, voxel-based fiber tracing maps are obtained for each subregion. For each sample, all voxels corresponding to each fiber bundle are extracted based on a mask, and their averages are used to obtain the connectivity fingerprint D between each subregion of the region of interest and each fiber bundle at the individual level. k×l×s The number of subregions is k, the number of fiber bundles is l, and the number of individuals is s;

[0030] C) The fingerprint of the connection between the individual-level region of interest (ROI) subregion and each fiber bundle. k×l×s Averaging was performed to obtain the connectivity fingerprints E between each subregion and subcortical fiber bundles based on the group. k×l The number of subregions is k, and the number of fiber bundles is l;

[0031] D) The connection fingerprint of the region of interest (ROI) subregion and each fiber bundle at the individual level. k×l× Group-level screening was performed, and fiber bundles with effective connections in different species were selected as features. Based on the screening results, the connection fingerprint E' between each subregion and the fiber bundle was extracted. k×l The number of subregions is k, and the number of fiber bundles is l.

[0032] Furthermore, in the aforementioned cross-species homology comparison, the connectivity fingerprints C' between each subregion of the obtained region of interest (ROI) and the gray matter brain region are analyzed. k×j and the connection fingerprint E' with the fiber bundle k×l As feature input, similarity or difference measures (including but not limited to Euclidean distance, Manhattan distance, cosine distance and KL divergence) between different subregions of different species are calculated. After normalizing the Euclidean distance, Manhattan distance, cosine distance and KL divergence respectively, the sum of the reciprocals of the four measures is taken as the score, and the score is used as the homology comparison criterion.

[0033] Furthermore, in the cross-species homology verification, the normalized whole-brain time series is obtained based on fMRI functional images, the time series corresponding to each subregion of the region of interest are extracted, and the functional connectivity from each subregion of different species to the whole brain is calculated. The whole-brain functional connectivity patterns of homologous subregions are compared, and the consistency of functional connectivity patterns between species is used as a verification of the reliability of this method. Attached Figure Description

[0034] The accompanying drawings, which are included to provide a further understanding of this application and constitute a part of this application, will be briefly described below to illustrate the technical solutions of the present invention more clearly.

[0035] Figure 1 This is a flowchart illustrating a cross-species comparison method for brain region homology based on gray matter brain regions and fiber bundles, as proposed in an embodiment of the present invention.

[0036] Figure 2 This is a schematic diagram of the process for extracting and dividing brain regions of interest according to an embodiment of the present invention;

[0037] Figure 3 This is a schematic flowchart illustrating the process of calculating the connection fingerprint between a subregion of interest and a gray matter brain region, as proposed in an embodiment of the present invention.

[0038] Figure 4 This is a schematic diagram of the process for calculating the connection between the subregion of interest and the subcortical fiber bundle in the fingerprint, as proposed in an embodiment of the present invention. Detailed Implementation

[0039] The technical solutions of the present invention will now be clearly and completely described with reference to the accompanying drawings in the embodiments of the present invention.

[0040] Unless otherwise defined, the technical or scientific terms used in this application shall have the general meaning understood by a person skilled in the art to which this application pertains. Words such as “a,” “an,” “an,” “the,” “the,” and “these” used in this application do not indicate quantitative limitation and may be singular or plural. The terms “comprising,” “including,” “having,” and any variations thereof used in this application are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or device that comprises a series of steps or modules (units) is not limited to the listed steps or modules (units) but may include steps or modules (units) not listed, or may include other steps or modules (units) inherent to these processes, methods, products, or devices. “Multiple” used in this application refers to two or more. “And / or” describes the relationship between related objects, indicating that three relationships may exist; for example, “A and / or B” can represent: A alone, A and B simultaneously, and B alone. Generally, the character “ / ” indicates that the objects before and after it are in an “or” relationship. The terms “first,” “second,” “third,” etc., used in this application are merely for distinguishing similar objects and do not represent a specific order of the objects.

[0041] Overall, this embodiment provides a cross-species comparison method for brain region homology based on gray matter brain regions and fiber tracts. Figure 1This is a flowchart of the cross-species comparison method for brain region homology based on gray matter brain regions and fiber tracts in this embodiment. The process includes: data acquisition and preprocessing, region of interest (ROI) extraction and brain region segmentation, connectivity fingerprint calculation between ROI subregions and gray matter brain regions, connectivity fingerprint calculation between ROI subregions and subcortical fiber tracts, cross-species homology comparison, and homology verification. For data acquisition and preprocessing, this example uses standard human and chimpanzee brain magnetic resonance imaging (MRI) data, including structural T1w images, functional fMRI images, and diffusion-weighted imaging (DTI) images. The T1w, fMRI, and DTI images are preprocessed separately. For ROI extraction and segmentation, corresponding ROIs are extracted from standard human and chimpanzee brain templates and segmented to obtain the segmentation results for the two species. The computation of connectivity fingerprints between subregions of interest (ROIs) and gray matter brain regions was performed using probabilistic fiber tracing based on the whole brain. After standardization, denoising, and false positive removal, structural connectivity from ROIs to the whole brain was obtained for both species. Individual-level connectivity fingerprints between subregions and gray matter brain regions were then group-level filtered to extract valid connectivity fingerprints from each subregion to the gray matter brain region. The computation of connectivity fingerprints between ROIs and subcortical fiber tracts was performed by reconstructing each fiber tract in different species using probabilistic fiber tracing, resulting in a voxel-based fiber tracing map for each subregion. The connectivity probability between each ROI subregion and each fiber tract was obtained by averaging all voxels in each fiber tract. Fiber tracts with valid connections in both species were selected as features, and valid connectivity fingerprints between each subregion and its fiber tract were extracted. Cross-species homology comparison used the connectivity fingerprints between each subregion and gray matter brain region and the connectivity fingerprints between each subregion and its fiber tract as feature inputs to calculate similarity or difference measures between human and chimpanzee subregions. The similarity measures between each ROI subregion of the two species were scored.

[0042] The above-described cross-species comparison method based on gray matter brain regions and fiber tracts for brain region homology involves preprocessing structural, functional, and diffusion-weighted data from multimodal brain imaging data of two species according to data acquisition and preprocessing. Regions of interest (ROIs) are extracted and segmented to obtain ROIs for both species. Then, connectivity fingerprints between ROI subregions and gray matter brain regions are calculated. Probabilistic fiber tracing is used to obtain connectivity fingerprints from each ROI subregion to cortical brain regions and subcortical nuclei. These fingerprints are then filtered to obtain connectivity fingerprints between ROI subregions and gray matter brain regions. Next, connectivity fingerprints between ROI subregions and subcortical fiber tracts are calculated to obtain connectivity fingerprints between each ROI subregion and fiber tract. Finally, cross-species homology scores for each ROI subregion are calculated using cross-species homology comparison. This method achieves cross-species comparison of gray matter brain regions and fiber tracts, providing a new approach and method for cross-species brain region homology comparison.

[0043] In some of these embodiments, the brain MRI data from different species are sourced from: brain MRI data from 40 healthy humans and brain MRI data from 46 healthy chimpanzees, including structural T1 data, functional fMRI data, and diffusion-weighted TTI data.

[0044] In some of these embodiments, a method for extracting regions of interest and segmenting brain regions is provided. Figure 2 This is a flowchart of the brain region extraction and segmentation method of this embodiment, including:

[0045] Region of interest extraction: Identify the region of interest to be studied and extract the corresponding region from the standard brain template of human and chimpanzee T1w structural images respectively.

[0046] The extracted region of interest (ROI) is registered and aligned with the diffusion space. The region of interest (ROI) of the extracted T1w structural image is registered and aligned with the diffusion space of the diffusion-weighted image. The region of interest (ROI) of the diffusion space is created as a template mask for the seed point. For each voxel in the mask, probabilistic fiber tracing is used to estimate the connection probability, and the connection matrix M between each voxel and the whole brain voxels is obtained.

[0047] Downsampling of the whole-brain connectivity matrix at each voxel in the region of interest (ROI) forms the voxel connectivity matrix A within the ROI. m×n The number of voxels in the ROI is m, and the number of voxels in the whole brain is n;

[0048] Calculate the cross-correlation matrix A between the whole-brain connectivity matrices of all voxels in the mask. m×m And used for automatic brain region segmentation;

[0049] Clustering was used on the above cross-correlation matrix to group voxels with higher similarity into 2-10 groups and the silhouette coefficient (SC) and sum of squared errors (SSE) of the clustering index for different groups were calculated. The k value with the highest clustering index score was selected as the division of the region of interest in different species, and the k subregion division results of the region of interest in different species were obtained respectively.

[0050] In some embodiments, a method for calculating the connectivity fingerprint between a region of interest subregion and a gray matter brain region is provided. Figure 3 This is a flowchart of the fingerprint calculation method for the connectivity between the region of interest subregion and the gray matter brain region in this embodiment, including:

[0051] Homology mapping was constructed by extracting brain atlases based on the MNI standard space template. The Desikan-KillianyAtlas atlas was used for the human brain, and the CY29 template for the chimpanzee brain. Based on the correspondence between these two templates in different species, homology maps were constructed for each species, corresponding one-to-one with the existing standard space brain templates.

[0052] Structural connectivity from subregions of interest (ROI) to the whole brain was constructed. Whole-brain probabilistic fiber tracing was used for each subregion of ROI in both humans and chimpanzees. Standardization, noise reduction, and false positive removal were performed to obtain the structural connectivity from each subregion of ROI to the whole brain for each sample from different species. k×n×s The number of subregions is k, the number of whole-brain voxels is n, and the number of individuals is s.

[0053] Fingerprint calculation of connections from subregions of interest to gray matter brain regions, and connection of each subregion to the whole brain structure B. k×n×s Using homology maps as templates, the average structural connectivity from an individual's subregion to the gray matter brain region is taken as the connectivity strength from the subregion of the region of interest (ROI) to that brain region, thus obtaining the individual-level connectivity fingerprint B from the subregion of the ROI to the cortical brain region and subcortical nuclei. k×j×s For the above-mentioned individual-level connection fingerprints B k×j×s Group-level averaging was performed to obtain the connectivity fingerprint between group-based region of interest (ROI) subregions and gray matter brain regions. k×j The number of subregions is k, the number of gray matter brain regions is j, and the number of individuals is s.

[0054] Group-level screening, at the group level, analyzed the connectivity fingerprints of individual cortical brain regions and subcortical nuclei in both humans and chimpanzees. k×j×s Statistical analysis was conducted on the individual-level connection fingerprints B. k×j×s Perform binarization processing on the binarized B k×j×s The average value of the connectivity fingerprints at the group level is taken to obtain the group probability. Connectivity fingerprints with a species group probability greater than 0.6 are selected as features.

[0055] Effective connectivity feature extraction: Based on the selected cortical region connectivity fingerprints, connectivity fingerprints C' from each subregion of the region of interest to the gray matter region are extracted. k×j For subsequent analysis, the number of subregions is k, and the number of gray matter brain regions is j:

[0056] In some of these embodiments, a method for calculating the probability of subcortical fiber bundle connections is provided. Figure 4 This is a flowchart of the method for calculating the probability of subcortical fiber bundle connectivity in this embodiment, including:

[0057] Shared fiber bundles are constructed. Based on the fiber bundles shared by different species as the basis for cross-species comparison, each fiber bundle is reconstructed for different species. After constructing a mask for each fiber bundle in the MNI space, it is registered to the individual to obtain the mask of the individual fiber bundle.

[0058] Connections from subregions of interest to common fiber bundles were constructed. Whole-brain probabilistic fiber tracing was used for each subregion of interest in both humans and chimpanzees, with a streamline count of 5000 for both species. False positives were removed using a 0.05% threshold, resulting in voxel-based fiber tracing maps for each subregion. For each sample in each species, all voxels corresponding to each fiber bundle were extracted based on a mask, and their average was used to obtain the connection fingerprint D between each subregion of interest and each fiber bundle. k×l×s Given k subregions, l fiber bundles, and s individuals, the fingerprint D is generated by connecting the fiber bundles at the individual level. k×l×s Averaging is performed to obtain the connection fingerprint E of subregions and fiber bundles based on groups. k×l The number of subregions is k, and the number of fiber bundles is l.

[0059] Group-level screening, at the group level, identifies the connectivity fingerprints (D) from each subregion of the region of interest to the common fiber bundle in both humans and chimpanzees. k×l×s Statistical analysis was conducted on the connection fingerprints at the individual level. k×l×s Perform binarization processing on the binarized D k×j×s The average value of the connectivity fingerprints at the group level is taken to obtain the group probability. Connectivity fingerprints with a species group probability greater than 0.6 are selected as features.

[0060] Effective connectivity feature extraction: Based on the fiber bundle connectivity fingerprints selected above, connectivity fingerprints E' from each subregion of the region of interest to the fiber bundle are extracted based on the group. k×l For subsequent analysis, the number of subregions is k, and the number of fiber bundles is l.

[0061] In some embodiments, a method for comparing cross-species homology is provided. The specific comparison criteria include: calculating the Euclidean distance, Manhattan distance, cosine distance, and KL divergence between each subregion of the region of interest of different species; normalizing the Euclidean distance, Manhattan distance, cosine distance, and KL divergence; and taking the sum of the reciprocals of the four metrics as the score. The smaller the distance, the higher the score, indicating that a certain subregion of one species is more homologous to a certain subregion of another species.

[0062] In some of the embodiments, a cross-species homology verification method is provided, which specifically includes: obtaining normalized whole-brain time series from preprocessed fMRI functional images of humans and chimpanzees, extracting time series corresponding to each subregion of the region of interest, calculating functional connectivity from each subregion of different species to the whole brain, comparing the whole-brain functional connectivity patterns of homologous subregions, and verifying the reliability of the method by whether the functional connectivity patterns between species are consistent.

[0063] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A cross-species comparative method for brain region homology based on gray matter brain regions and fiber tracts, characterized in that, Includes the following steps: S1. Obtain standard brain magnetic resonance data of different species and preprocess them to obtain preprocessed structural image data, functional image data and diffusion-weighted image data; S2. Extract the corresponding regions of interest (ROI) from the standard brain templates of different species and apply probabilistic fiber tracing and clustering methods to divide the region into subregions of the ROI. S3. Construct homology maps of cortical brain regions and subcortical nuclei in different species, apply probabilistic fiber tracing to obtain structural connections from each subregion of the region of interest (ROI) to the whole brain, screen effective connections within species at the group level, and extract the connection fingerprints between each subregion of the ROI and the gray matter brain region. S4. Extract fiber bundles common to different species as templates, obtain voxel-based fiber tracking maps of each subregion of the region of interest (ROI) based on probabilistic fiber tracking, calculate the connection probability value between each subregion of the ROI and each fiber bundle, statistically analyze the fiber bundle probability value within the species at the group level, and extract the connection fingerprint between each subregion of the ROI and the fiber bundle. S5. Using the connectivity fingerprint between each subregion of the region of interest (ROI) and the gray matter brain region and fiber bundle as feature input, calculate the similarity or difference between each subregion of different species and score it. The score is used as the criterion for homology comparison. Functional connectivity from each subregion to the whole brain is calculated based on functional image data, and the consistency of whole-brain functional connectivity patterns of homologous subregions in different species is used as a verification of the reliability of this method.

2. The cross-species comparison method for brain region homology based on gray matter brain regions and fiber tracts according to claim 1, characterized in that, The process involves acquiring and preprocessing standard brain magnetic resonance imaging (MRI) data from different species, including structural T1w images, functional fMRI images, and diffusion-weighted tensor (DTI) images. Specifically, the preprocessing includes denoising, registration, skull removal, and brain tissue segmentation for the T1w structural images; motion correction, spatial standardization, smoothing, signal extraction, and normalization for the fMRI functional images; and denoising, registration, correction, brain tissue extraction, and diffusion tensor fitting for the DTI images.

3. The method for cross-species comparison of brain region homology based on gray matter brain regions and fiber tracts according to claim 2, characterized in that, Step S2 includes: S21. Region of Interest (ROI) Extraction: Identify the region of interest (ROI) to be studied and extract the corresponding region from the standard brain template of T1w structural images of different species. S22. Based on the preprocessed structural image data and diffusion-weighted image data, the region of interest (ROI) of the extracted T1w structural image is registered and aligned with the diffusion space of the diffusion-weighted image. The region of interest (ROI) of the diffusion space is created as a template mask for the seed point. For each voxel in the mask, probabilistic fiber tracing is used to estimate the connection probability, and the connection matrix M between each voxel and the whole brain voxels is obtained. S23. Downsample the whole-brain connectivity matrix M at each voxel in the region of interest (ROI) to form the voxel connectivity matrix A within the ROI. m×n The number of voxels in the ROI is m, and the number of voxels in the whole brain is n; S24. Calculate the cross-correlation matrix A between the whole-brain connectivity matrices of all voxels within the region of interest (ROI). m×m And used for automatic brain region segmentation; S25. Apply a clustering method to the cross-correlation matrix, grouping voxels with higher similarity together to form different groups and calculating the clustering index of these groups. Select the optimal solution k of the clustering index as the division of the region of interest (ROI) in different species, and obtain the k division results of the region of interest (ROI) for different species respectively.

4. The cross-species comparison method for brain region homology based on gray matter brain regions and fiber tracts according to claim 3, characterized in that, Step S3 includes: S31. Based on the existing T1w structural image standard brain atlas, construct homology maps of gray matter brain regions in different species, including cortical brain regions and subcortical nuclei. S32. Based on the preprocessed structural image data and diffusion-weighted image data, a probabilistic fiber tracing algorithm based on the whole brain is applied to each subregion of different species. The fiber tracing results are then standardized, denoised, and false positives are removed to obtain the structural connectivity B from different species subregions to the whole brain. k×n×s The number of subregions is k, the number of whole-brain voxels is n, and the number of individuals is s; S33, Connecting different species subregions to the whole brain (B) k×n×s Using a homology map as a template, and taking voxels within each gray matter region of the homology map as a whole, the average structural connectivity from an individual's subregion to the gray matter region is taken as the connectivity strength from the subregion of the region of interest (ROI) to that region. This yields the connectivity fingerprint B from the subregion of the ROI to the cortical region and subcortical nuclei at the individual level. k×j×s The number of subregions is k, the number of gray matter brain regions is j, and the number of individuals is s; S34. Connectivity fingerprints from the individual-level region of interest (ROI) subregions to cortical brain regions and subcortical nuclei B k×j×s Group-level averaging was performed to obtain the connectivity fingerprint between group-based region of interest (ROI) subregions and gray matter brain regions. k×j The number of subregions is k, and the number of gray matter brain regions is j; S35, Connectivity fingerprints of cortical regions and subcortical nuclei at the individual level (B) k×j×s Group-level screening was performed to obtain effective intraspecific connections as features. Based on the screening results, connection fingerprints C' between group-based region of interest (ROI) subregions and gray matter brain regions were extracted. k×j The number of subregions is k, and the number of gray matter brain regions is j.

5. The cross-species comparison method for brain region homology based on gray matter brain regions and fiber tracts according to claim 4, characterized in that, Step S4 includes: S41. Based on the fiber bundles common to different species as the basis for cross-species comparison, each fiber bundle is reconstructed for different species. After constructing a mask for each fiber bundle in the MNI space, it is registered to the individual to obtain the mask of the individual fiber bundle. S42. Using probabilistic fiber tracing, a voxel-based fiber tracing map is obtained for each subregion. For each sample, all voxels corresponding to each fiber bundle are extracted based on a mask, and their averages are used to obtain the connection fingerprint D between the region of interest (ROI) subregion and each fiber bundle at the individual level. k×l×s The number of subregions is k, the number of fiber bundles is l, and the number of individuals is s; S43. The connection fingerprint D between the individual-level region of interest (ROI) subregion and each fiber bundle k×l× S is averaged to obtain the group-based subregion and subcortical fiber bundle connectivity fingerprint E. k×l The number of subregions is k, and the number of fiber bundles is l; S44. The connection fingerprint D between the subregion of the region of interest (ROI) at the individual level and each fiber bundle. k×l×s Group-level screening was performed, selecting fiber bundles with effective connections across different species as features. Based on the screening results, the connection fingerprint E' between each subregion and the fiber bundle was extracted. k×l The number of subregions is k, and the number of fiber bundles is l.

6. The method for cross-species comparison of brain region homology based on gray matter brain regions and fiber tracts according to claim 5, characterized in that, At the group level, the connectivity fingerprints of each subregion of the Region of Interest (ROI) for different species were statistically analyzed. The connectivity fingerprints at the individual level were binarized, and the average value of the connectivity fingerprints at the group level was taken to obtain the group probability. Connectivity fingerprints with a species group probability greater than 0.6 were selected as features.

7. The method for cross-species comparison of brain region homology based on gray matter brain regions and fiber tracts according to claim 6, characterized in that, Cross-species homology comparison criteria were used to identify the connectivity fingerprints (C') between each subregion of the region of interest (ROI) and the gray matter brain regions. k×j and the connection fingerprint E' with the fiber bundle k×l As a feature input, the similarity or difference between different subregions of different species is calculated and scored, and the score is used as a criterion for homology comparison.

8. The cross-species comparison method for brain region homology based on gray matter brain regions and fiber tracts according to claim 7, characterized in that, To verify cross-species homology, functional connectivity from various subregions to the whole brain of different species is calculated based on functional image data. The functional connectivity patterns of homologous subregions in different species are compared to verify the reliability of this method.

Images