Automatic partitioning method and system for multi-modal brain atlas in combination with human connection group plan

Abstract

The invention discloses an automatic partitioning method and system for a multi-modal brain atlas in combination with a human connection group plan. The method comprises the following steps: firstly, acquiring T1 weighted sMRI data and fMRI data collected for a target tested brain, and converting the T1 weighted sMRI data and the fMRI data into an NIFTI format conforming to a BIDS standard; preprocessing and whole-brain segmentation reconstruction are carried out on the sMRI data, the preprocessed fMRI data are registered to a cortex surface model obtained through reconstruction, and the two types of registered data are converted to an MNI standard space; and finally, fMRI data in an MNI standard space is registered to a Conte69 template and then down-sampled to a spatial resolution required by a CIFTI space, and the two types of data are mapped to a CIFTI gray scale space and packaged. According to the method framework, mapping from a non-HCP standard protocol data set to an HCPMMP standard partition is realized for the first time, and possibility is provided for brain function research on a wider multi-mode data set.

Description

Technical Field

[0001] The present invention belongs to the field of brain imaging, and specifically relates to proposing an automatic brain parcellation method based on the Human Connectome Project Multi-Modal Parcellation (HCPMMP) based on a joint approach. Background Art

[0002] Brain parcellation is a key issue in neuroscience research. Accurate parcellation is of great significance for analyzing brain functional networks, revealing the mechanisms of brain diseases, and promoting the development of brain-computer interface technology. With the rapid advancement of high-resolution imaging technology, massive amounts of brain imaging data are constantly accumulating, providing unprecedented opportunities for achieving refined brain parcellation. HCPMMP, a multimodal brain parcellation scheme based on the Human Connectome Project, integrates multiple brain imaging data (including structural and functional information) to accurately divide the cerebral cortex into 180 bilateral regions (a total of 360 regions), thereby generating the most accurate and refined brain functional map currently available. This method provides a reliable tool for studying brain functional networks, personalized medicine, and diagnosing brain diseases.

[0003] However, the practical application of HCPMMP faces numerous challenges. Its parcellation process relies on MRI scan data that adheres to the strict HCP (Human Connectome Project) protocol. In addition to traditional structural T1-weighted and functional data, it also requires high-resolution structural T2-weighted data, seven task-specific functional data, diffusion MRI, and magnetic field distribution information. These demanding imaging conditions make it unsuitable for most neuroimaging datasets, severely limiting the widespread application of HCPMMP. Summary of the Invention

[0004] The purpose of the present invention is to address the limitations of the existing HCPMMP method in terms of data dependence, partitioning accuracy and scope of application, and to provide an automatic partitioning method for the multimodal brain map of the joint human connectome project.

[0005] The specific technical solutions adopted in the present invention are as follows:

[0006] In a first aspect, the present invention provides a method for automatically partitioning a multimodal brain map using a joint human connectome project, comprising:

[0007] S1. Obtain T1-weighted sMRI data and fMRI data collected from the target subject's brain, and the file formats of both modal data are DICOM format;

[0008] S2. Convert T1-weighted sMRI data and fMRI data from DICOM format to NIFTI format that complies with the BIDS standard and perform BIDS integrity check;

[0009] S3. Preprocess and reconstruct the whole brain segmentation of T1-weighted sMRI data that have passed BIDS integrity check to obtain the cortical surface model and subcutaneous structure;

[0010] S4, preprocessing the verified fMRI data, registering the preprocessed fMRI data to a cortical surface model reconstructed based on the T1-weighted sMRI data, and then converting the registered fMRI data and the T1-weighted sMRI data to MNI standard space;

[0011] S5. Register the fMRI data in the MNI standard space to the Conte69 template, then downsample to the spatial resolution required by the CIFTI space. Then map the T1-weighted sMRI data in the MNI standard space and the downsampled fMRI data to the CIFTI grayscale space, thereby integrating the two in a unified coordinate system. Finally, encapsulate the T1-weighted sMRI and fMRI data in coordinate system 1 into a data format that meets the HCP standard.

[0012] As a preferred embodiment of the first aspect, in S1, magnetic field distribution information may be further obtained for correcting image distortion caused by magnetic field inhomogeneity when preprocessing the verified fMRI data.

[0013] As a preferred embodiment of the first aspect, in S2, BIDS Validator is used to perform BIDS integrity verification on the T1-weighted sMRI data and fMRI data in NIFTI format, and subsequent steps can be executed only after the verification passes.

[0014] As a preferred embodiment of the first aspect, in S3, the T1-weighted sMRI data are preprocessed and the whole-brain segmentation and reconstruction are performed using commands integrated in the FreeSurfer tool.

[0015] As a preferred embodiment of the above-mentioned first aspect, in S4, the fMRIprep framework is used to preprocess and align the fMRI data, wherein the preprocessing operations include fMRI sequence slice calibration, rigid body calibration and magnetic susceptibility distortion correction, and the registration algorithm is the bbregister algorithm; at the same time, in S4, the ANTs software is used to convert the aligned fMRI data and T1-weighted sMRI data into the MNI standard space.

[0016] As a preferred embodiment of the first aspect, in S5, the T1-weighted sMRI and fMRI data are encapsulated into a data format that complies with the HCP standard using CIFTIFY (Connectivity Informatics Technology Initiative Format Analysis) technology.

[0017] In a second aspect, the present invention provides an automatic partitioning system for a multimodal brain atlas using a joint human connectome project, comprising:

[0018] The data acquisition module is used to obtain T1-weighted sMRI data and fMRI data collected from the brain of the target subject, and the file format of both modal data is DICOM format;

[0019] The conversion and verification module is used to convert T1-weighted sMRI data and fMRI data from DICOM format to NIFTI format that complies with the BIDS standard and perform BIDS integrity verification;

[0020] The structural image processing module is used to preprocess and reconstruct the whole brain segmentation of T1-weighted sMRI data that have passed the BIDS integrity check to obtain the cortical surface model and subcutaneous structure;

[0021] a functional image processing module, configured to preprocess the verified fMRI data, register the preprocessed fMRI data to a cortical surface model reconstructed based on the T1-weighted sMRI data, and then convert the registered fMRI data and T1-weighted sMRI data to the MNI standard space;

[0022] The registration and encapsulation module is used to register the fMRI data in the MNI standard space to the Conte69 template, then downsample it to the spatial resolution required by the CIFTI space, and then map the T1-weighted sMRI data in the MNI standard space and the downsampled fMRI data to the CIFTI grayscale space, thereby integrating the two in a unified coordinate system. Finally, the T1-weighted sMRI and fMRI data in coordinate system 1 are encapsulated into a data format that conforms to the HCP standard.

[0023] In a third aspect, the present invention provides a computer program product comprising a computer program / instruction, which, when executed by a processor, can implement the method for automatically partitioning the multimodal brain map of the Joint Human Connectome Project as described in any one of the schemes of the first aspect above.

[0024] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for automatically partitioning the multimodal brain map of the joint human connectome project as described in any one of the schemes in the first aspect above can be implemented.

[0025] In a fifth aspect, the present invention provides a computer electronic device comprising a memory and a processor;

[0026] The memory is used to store computer programs;

[0027] The processor is used to implement the automatic partitioning method of the joint human connectome project multimodal brain map as described in any one of the solutions of the first aspect above when executing the computer program.

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] This paper proposes a combined HCPMMP automatic brain parcellation method. This framework requires only T1-weighted structural Magnetic Resonance Imaging (sMRI) and functional Magnetic Resonance Imaging (fMRI) to achieve precise automatic parcellation of the cerebral cortex, effectively reducing its reliance on specific high-standard imaging protocols, thereby expanding its applicability in diverse data environments. This framework, for the first time, achieves the mapping of non-HCPMMP standard protocol datasets to HCPMMP standard parcellations, breaking through HCPMMP's high reliance on dataset standards and facilitating brain function research on a wider range of multimodal datasets. This invention not only significantly enhances the practicality and versatility of the HCPMMP parcellation method but also lays a solid foundation for multimodal automatic identification of cognitive impairments based on complex brain network theory. This invention is expected to significantly promote the in-depth utilization and mining of brain imaging data, providing important technical support for precision medicine and intelligent healthcare, and possesses extremely high scientific value and application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 Flowchart of the steps for the automatic partitioning method based on the Joint Human Connectome Project multimodal brain atlas;

[0031] Figure 2 Schematic diagram comparing the brain partitioning framework of the traditional method and the method of the present invention;

[0032] Figure 3 Schematic diagram of the modular composition of the automatic partitioning system for the joint human connectome project multimodal brain atlas;

[0033] Figure 4 It is a schematic diagram of the structure of computer electronic equipment;

[0034] Figure 5 A schematic diagram of some results of an embodiment of the present invention;

[0035] Figure 6 This is an accurate brain partitioning map of clinical data according to an embodiment of the present invention. DETAILED DESCRIPTION

[0036] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings. In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways than those described herein, and those skilled in the art can make similar improvements without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below. The technical features in the various embodiments of the present invention can be combined accordingly without conflicting with each other.

[0037] like Figure 1 As shown, in a preferred embodiment of the present invention, a method for automatically partitioning a multimodal brain map in conjunction with a human connectome project is provided, characterized in that it includes:

[0038] S1. Obtain T1-weighted sMRI data and fMRI data collected from the target subject's brain, and the file format of both modality data is DICOM format.

[0039] It should be noted that T1-weighted sMRI data and fMRI data are two necessary data modalities that need to be collected in the present invention. However, if more data modalities can be collected, the advantages of the present invention in multimodal fusion can be more fully utilized, thereby obtaining a more accurate brain function map.

[0040] S2. Convert T1-weighted sMRI data and fMRI data from DICOM format to NIFTI format that complies with the BIDS standard and perform BIDS integrity check.

[0041] It should be noted that BIDS (Brain Imaging Data Structure) is a standardized organizational framework designed for neuroimaging data, aiming to address the challenges of data sharing and reproducible research in the field of neuroscience. The standard was originally developed for MRI data and has now been extended to multiple modalities such as EEG, MEG, PET, and supports the integration of behavioral data and physiological data. BIDS has significantly improved the accessibility, reproducibility and interoperability of neuroimaging data by standardizing data organization, naming and metadata descriptions, and has become an infrastructure-level standard in the field of neuroscience. The specific data specifications of the BIDS standard belong to the prior art and will not be repeated in this invention.

[0042] The original data formats of the above-mentioned T1-weighted sMRI data and fMRI data are both in DICOM format. In order to facilitate subsequent data processing, they need to be converted into NIFTI format that conforms to the BIDS standard. This data format conversion can be achieved by designing program code by yourself, or by combining it with existing software programs such as dcm2niix. At the same time, BIDS integrity check is required after format conversion to ensure that the converted file conforms to the BIDS standard.

[0043] In an embodiment of the present invention, in step S2, the BIDS Validator is used to perform a BIDS integrity check on the T1-weighted sMRI and fMRI data in NIFTI format. Only after the check passes can the subsequent steps of the process be executed. If the check fails, the process must be terminated and the data conversion must be repeated. The BIDS Validator is an existing, mature tool that can be directly called to perform the BIDS integrity check of the present invention. Of course, in addition to the BIDS Validator, a custom program can also be used to convert the DICOM format to a NIFTI format that complies with the BIDS standard, as long as the format conversion and BIDS integrity check can be achieved.

[0044] S3. Preprocess and reconstruct the whole brain segmentation of the T1-weighted sMRI data that have passed the BIDS integrity check to obtain the cortical surface model and subcutaneous structure.

[0045] In an embodiment of the present invention, the commands integrated in the FreeSurfer tool are used to preprocess the T1-weighted sMRI data and perform whole-brain segmentation and reconstruction. The FreeSurfer tool is an open source neuroimaging analysis tool that focuses on processing and analyzing human brain magnetic resonance imaging (MRI) data. It provides a complete set of processing procedures, including skull stripping, B1 bias field correction, gray and white matter segmentation, cortical surface reconstruction, etc., which can generate high-precision three-dimensional brain models and calculate anatomical parameters such as cortical thickness, surface area, and gray matter volume. In an embodiment of the present invention, the command recon_all integrated in the FreeSurfer tool can be directly used to implement the above-mentioned S3 step. recon-all is the core automated processing pipeline instruction in the FreeSurfer toolkit, which is used to perform full-process analysis of structural magnetic resonance imaging (sMRI) data and generate a variety of anatomical quantitative indicators of the cerebral cortex surface model and subcutaneous structures.

[0046] S4. Preprocess the fMRI data that has passed the BIDS integrity check, and align the preprocessed fMRI data to the cortical surface model reconstructed based on the T1-weighted sMRI data, and then convert the aligned fMRI data and T1-weighted sMRI data into the MNI standard space.

[0047] In an embodiment of the present invention, in step S4, the fMRIprep framework can be used to preprocess and register the fMRI data. The preprocessing operations include fMRI sequence slice alignment, rigid body alignment, and susceptibility distortion correction. The registration criterion uses the cortical surface model obtained in step S3 as a reference, and the fMRI data is registered to the cortical surface model using the bbregister algorithm. These preprocessing and registration operations can be directly implemented by calling instructions integrated into the fMRIprep framework. These are standard operations in MRI data processing and will not be further described.

[0048] In an embodiment of the present invention, in the above step S4, the ANTs software can be used to convert the registered fMRI data and T1-weighted sMRI data into the MNI standard space. ANTs (Advanced Normalization Tools) is an open source software for medical image analysis. Its core functions include linear registration (such as rigid body, affine transformation) and nonlinear registration (such as SyN algorithm), which can be used to convert individual data into a standard space. In the present invention, the built-in registration tool of ANTs can be used to register the registered fMRI data and T1-weighted sMRI data to the MNI template, so that the two types of MRI data are aligned and converted into the MNI standard space.

[0049] S5. Register the fMRI data in the MNI standard space to the Conte69 template, then downsample to the spatial resolution required by the CIFTI space. Then map the T1-weighted sMRI data in the MNI standard space and the downsampled fMRI data to the CIFTI grayscale space, thereby integrating the two in a unified coordinate system. Finally, encapsulate the T1-weighted sMRI and fMRI data in coordinate system 1 into a data format that meets the HCP standard.

[0050] In an embodiment of the present invention, in step S5 above, CIFTIFY (Connectivity Informatics Technology Initiative Format Analysis) technology is used to encapsulate T1-weighted sMRI and fMRI data into a data format that complies with the HCP standard. CIFTIFY is an existing, mature toolkit that can convert the formats of FreeSurfer reconstruction results and fMRI data. The CIFTI format supports the integration of surface and voxel data, facilitating cross-subject and multimodal analysis. Users can use the command-line tools provided by CIFTIFY to perform data conversion. Therefore, the specific conversion and encapsulation methods are state-of-the-art and will not be described in detail.

[0051] like Figure 2Figure 1 shows a comparison between a traditional HCP minimization pipeline and the method of the present invention. In the traditional HCP minimization pipeline, the parcellation process relies on an HCP protocol dataset that adheres to a strict HCP protocol, including T1-weighted (i.e., T1w) data, T2-weighted (i.e., T2w) data, resting-state functional magnetic resonance imaging (fMRI) data, magnetic field distribution data, and other data, all of which are essential. However, the method of the present invention, shown in S1-S5 above, does not rely on an HCP protocol dataset. Instead, it only requires fMRI data and T1-weighted sMRI data as the basic data modalities (with the optional addition of magnetic field distribution data) to obtain a data format that complies with the HCP standard and is applied to the HCP PMMP minimization pipeline. Therefore, the method of the present invention uses T1-weighted sMRI and fMRI as the minimum required data modalities, meaning that even datasets containing only basic structural and functional information can achieve reliable brain parcellation. Of course, this method framework is highly flexible and extensible. When additional modal data (such as high-resolution T2-weighted structural imaging, magnetic field distribution information, etc.) is provided, the accuracy and robustness of the partitioning results can be further improved. During the data acquisition process, it is recommended to collect sMRI and fMRI data measured at the same time interval to ensure spatiotemporal consistency. If available, other modal information should also be collected simultaneously as much as possible to give full play to the advantages of the method framework of the present invention in multimodal fusion, thereby obtaining a more accurate brain functional map.

[0052] In an embodiment of the present invention, in addition to the necessary T1-weighted sMRI and fMRI data, magnetic field distribution information may be further obtained in step S1 to correct image distortion caused by magnetic field inhomogeneity during preprocessing of the verified fMRI data (i.e., for the magnetic susceptibility distortion correction step in step S4). Of course, if only T1-weighted sMRI and fMRI data are available, but magnetic field distribution information is missing, then the magnetic susceptibility distortion correction step in step S4 can utilize device parameters stored in a NIFTI format file that complies with the BIDS standard for general distortion correction.

[0053] It should be noted that the method steps shown in S1 to S5 above can essentially be implemented in the form of computer programs or software function modules.

[0054] Therefore, based on the same inventive concept, Figure 3 As shown, the present invention also provides an automatic partitioning system for a multimodal brain map of a joint human connectome project, corresponding to the automatic partitioning method for a multimodal brain map of a joint human connectome project provided in the above embodiment, comprising:

[0055] The data acquisition module is used to obtain T1-weighted sMRI data and fMRI data collected from the brain of the target subject, and the file format of both modal data is DICOM format;

[0056] The conversion and verification module is used to convert T1-weighted sMRI data and fMRI data from DICOM format to NIFTI format that complies with the BIDS standard and perform BIDS integrity verification;

[0057] The structural image processing module is used to preprocess and reconstruct the whole brain segmentation of T1-weighted sMRI data that have passed the BIDS integrity check to obtain the cortical surface model and subcutaneous structure;

[0058] a functional image processing module, configured to preprocess the verified fMRI data, register the preprocessed fMRI data to a cortical surface model reconstructed based on the T1-weighted sMRI data, and then convert the registered fMRI data and T1-weighted sMRI data to the MNI standard space;

[0059] The registration and encapsulation module is used to register the fMRI data in the MNI standard space to the Conte69 template, then downsample it to the spatial resolution required by the CIFTI space, and then map the T1-weighted sMRI data in the MNI standard space and the downsampled fMRI data to the CIFTI grayscale space, thereby integrating the two in a unified coordinate system. Finally, the T1-weighted sMRI and fMRI data in coordinate system 1 are encapsulated into a data format that conforms to the HCP standard.

[0060] In addition, based on the same inventive concept, Figure 4 As shown, the present invention also provides a computer electronic device corresponding to the automatic partitioning method of the multimodal brain map of the joint human connectome project provided in the above embodiment, which includes a memory and a processor;

[0061] The memory is used to store computer programs;

[0062] The processor is configured to implement the aforementioned automatic partitioning method of the joint human connectome project multimodal brain map when executing the computer program.

[0063] Furthermore, the logic instructions in the aforementioned memory can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention.

[0064] Therefore, based on the same inventive concept, the present invention provides a computer-readable storage medium corresponding to the automatic partitioning method of the multimodal brain map of the Joint Human Connectome Project, and the storage medium stores a computer program. When the computer program is executed by the processor, it can implement the automatic partitioning method of the multimodal brain map of the Joint Human Connectome Project as described above.

[0065] Therefore, based on the same inventive concept, the present invention provides a computer program product, including a computer program / instruction, which, when executed by a processor, can implement the automatic partitioning method of the multimodal brain map of the joint human connection group project as described above.

[0066] Specifically, in the computer-readable storage medium of the above three embodiments, the stored computer program is executed by the processor to perform the above steps S1 to S5.

[0067] It is understood that the storage medium may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Furthermore, the storage medium may be any medium capable of storing program code, such as a USB flash drive, a mobile hard drive, a magnetic disk, or an optical disk.

[0068] It is understandable that the above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0069] It should also be noted that those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working process of the system described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here. In the various embodiments provided in this application, the division of steps or modules in the system and method is only a logical function division. In actual implementation, there may be other division methods, for example, multiple modules or steps can be combined or integrated together, and a module or step can also be split.

[0070] The present invention will further demonstrate the detailed implementation process and technical effects of the automatic partitioning method of the joint human connectome project multimodal brain map shown in the above steps S1 to S5 on specific example data through a specific embodiment, so as to facilitate understanding of the essence of the present invention.

[0071] Example

[0072] The steps of this embodiment are identical to the method for automatically partitioning the multimodal brain map of the Joint Human Connectome Project shown in steps S1-S5 above, so we will not repeat them here. This embodiment will focus on the specific data, some specific settings, and implementation results. For ease of description, the method for automatically partitioning the multimodal brain map of the Joint Human Connectome Project shown in steps S1-S5 above will be referred to as the present invention.

[0073] Step 1: Data Collection

[0074] Raw imaging data (DICOM format, maintaining spatiotemporal consistency) was collected from different patients. The data modalities included two basic modalities: T1-weighted structural magnetic resonance imaging (sMRI) data and functional magnetic resonance imaging (fMRI) data. In addition, magnetic field distribution information was also collected for some patient samples. The raw data was sorted into separate folders according to different modalities, and these folders were packaged and compressed into a compressed file named after the patient's ID number.

[0075] Step 2: BIDS converts and verifies the image

[0076] Differences in device parameters can lead to variations in the imaging quality of neuroimaging data. To effectively manage and verify this complex data, this example first employs a conversion algorithm based on the BIDS standard to perform format conversion and verify data integrity. BIDS (Brain Imaging Data Structure) was proposed by Gorgolewski et al. from Stanford University in 2016. It aims to standardize and organize neuroimaging data using a unified standard, thereby simplifying data management processes and improving data consistency and usability. This example uses software to convert the original acquired DICOM (Digital Imaging and Communications in Medicine) files into the NIFTI (Neuroimaging Informatics Technology Initiative) format, which complies with the BIDS standard. The data is then classified and managed based on information such as the subject's specific visit time and imaging modality type. Furthermore, during the BIDS conversion process, data quality is restored and standardized to a certain extent based on the imaging parameters of different devices, reducing deviations caused by device differences. This approach not only improves the compatibility and consistency of multicenter data but also facilitates longitudinal studies of lesions over multiple follow-up visits. It further promotes the integration and comparative analysis of multimodal data, providing reliable technical support for large-scale neuroimaging research.

[0077] In this embodiment, both T1-weighted sMRI data and fMRI data need to be converted from DICOM format to NIFTI format that complies with the BIDS standard and undergo BIDS integrity verification. Only after passing the verification can subsequent processes be carried out.

[0078] Step 3: Reconstruct the structural image

[0079] This example uses the FreeSurfer tool to preprocess and reconstruct whole-brain segmentation on T1-weighted sMRI data that have passed BIDS integrity verification. Specifically, FreeSurfer automatically executes the entire preprocessing and whole-brain segmentation reconstruction process by executing the recon-all command. This command accurately determines the boundaries of the cerebral cortex by analyzing the contrast between gray and white matter in T1-weighted images. It then extracts a series of key cortical morphological features, including "smoothness," "medium cortical thickness," "pial membrane," and "expansion," and obtains a cortical surface model. After processing, this morphological information is saved in the GIFTI (Generic Image File Format for Imaging Informatics) format to facilitate subsequent analysis and modeling. In addition to reconstructing the cortical surface model, subcutaneous structures are extracted, and the number and volume of voxels in each subcutaneous tissue are counted. Thus, through FreeSurfer's high-precision preprocessing, microstructural features of the brain can be accurately captured, providing reliable support for further brain parcellation and neuroimaging research.

[0080] Step 4: Functional image processing

[0081] This embodiment uses the fMRIprep framework to preprocess the fMRI data that has passed the BIDS integrity check, and registers the preprocessed fMRI data to the cortical surface model reconstructed based on the T1-weighted sMRI data.

[0082] The specific preprocessing process in this embodiment includes: first, using AFNI (Analysis of Functional Neuroimages) to calibrate the fMRI sequence slices, and align all slices to the middle position of each echo time through interpolation technology to ensure the temporal consistency between slices. Then, use FSL (FMRIB Software Library) software to perform six-parameter rigid body calibration of head movement to correct artifacts caused by movement and ensure the accuracy and stability of image data. In addition, this embodiment also combines the magnetic field distribution information collected in S1 to perform magnetic susceptibility distortion correction to reduce image distortion caused by magnetic field inhomogeneity. If the magnetic field distribution information is not collected in a patient sample, the device parameters stored in the BIDS structure will be used to perform general distortion correction to ensure that the data can still be effectively processed in the absence of certain specific information.

[0083] In the process of registering the pre-processed fMRI data to the cortical surface model, this embodiment uses the boundary-based bbregister algorithm to ensure accurate alignment of the two modalities, thereby achieving optimal fusion of structural and functional information.

[0084] Finally, in order to standardize the imaging data to a common space, this example uses ANTs software to convert the registered fMRI data and T1-weighted sMRI data to the MNI standard space, achieving conversion from the original sampling space to the MNI space and ensuring that the imaging results can be aligned with the brain standard template.

[0085] Through this series of processing steps in step 4, this embodiment can maximize the quality and consistency of fMRI data, providing reliable basic data for subsequent brain function research.

[0086] Step 5: Conte69 template registration

[0087] To implement HCPMMP partitioning and register fMRI data into the CIFTI space proposed by HCP, this example first registers fMRI data in the MNI standard space with the Conte69 template. The Conte69 template defines 163,842 (164K) vertices in the cerebral cortex grid for each hemisphere, representing the detailed spatial structure of the cerebral cortex. To conform the data to the requirements of the CIFTI space, this example downsamples these 164K vertex coordinates to the spatial resolution required by CIFTI space, ultimately obtaining 32,492 vertex coordinates, or standard CIFTI space coordinates (32K). This process ensures that the fMRI data can be represented in the HCP standard cortical grid and meets the requirements of HCPMMP partitioning.

[0088] Next, this example maps the structural and functional MRI data to the CIFTI grayscale space. Specifically, the MNI-standard T1-weighted sMRI data and the downsampled fMRI data are mapped to the CIFTI grayscale space, integrating them into a unified coordinate system. To ensure data uniformity and operability, this example uses CIFTIFY (Connectivity Informatics Technology Initiative Format Analysis) technology to encapsulate the T1-weighted sMRI and fMRI data in coordinate system 1 into a data format that complies with the HCP standard. This processing allows the data to be directly applied to the HCPMMP minimization pipeline.

[0089] In practical applications, the encapsulated data can be subsequently analyzed based on actual analysis needs. In this example, by further applying CIFTIFY technology, the fMRI data is partitioned to obtain time series information for each brain region. This process not only conforms to the HCP standard format but also provides high-quality input data for brain functional parcellation and network analysis, thereby supporting accurate brain parcellation and functional connectivity analysis.

[0090] The present invention can set various forms of result output according to the actual analysis requirements. In the embodiment of the present invention, the user can select the following two output modes according to specific needs. The operation results are shown in the following example: Figure 5 As shown:

[0091] Output mode 1, brain region time series data: output the average time series data of each brain region and save it in NIFTI format. The example results are as follows: Figure 5 As shown in (a) in the figure, this time series data can be used according to specific needs, such as using a sliding window for dynamic functional connectivity analysis.

[0092] Output mode 2, correlation matrix: Calculate the functional connectivity strength based on the output brain region time series data, generate the functional connectivity correlation matrix between each brain region, and save it in Mat format. The example results are as follows Figure 5 Data in this format can be directly read and analyzed using programming languages ​​such as Matlab or Python, providing direct data support for complex brain network analysis, functional connectivity pattern recognition, and pathological feature research.

[0093] Due to the large number of brain regions, only three regions' time series are displayed in the visualization of brain region time series data. This flexible output mode design provides researchers with a variety of analytical possibilities, meeting both simple visualization needs and supporting complex downstream analysis tasks. Through intuitive time series display and precise functional connectivity matrix calculation, the method of this invention has broad application prospects in functional network analysis and brain disease research.

[0094] In further experiments of this embodiment, Figure 6 The results shown demonstrate accurate brain parcellation maps obtained by the present method using automated parcellation of selected exemplary clinical data. A and B represent left brain parcellations, C and D represent right brain parcellations, and E and F are expanded schematic diagrams of cerebral cortex parcellation. This demonstrates that the present method, by appropriately reducing the constraints of the HCP scanning protocol, can directly process imaging data acquired by current clinical equipment, achieving multimodal parcellation of the human cortex consistent with HCPMMP.

[0095] The embodiments described above are merely some preferred implementations of the present invention and are not intended to limit the present invention. Persons skilled in the art may make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, any technical solution obtained by equivalent substitution or equivalent transformation falls within the scope of protection of the present invention.

Claims

1. An automatic partitioning method for multimodal brain maps in conjunction with the Human Connectome Project, characterized in that include: S1. Obtain T1-weighted sMRI data and fMRI data collected from the target subject's brain, and the file formats of both modal data are DICOM format; S2. Convert T1-weighted sMRI data and fMRI data from DICOM format to NIFTI format that complies with the BIDS standard and perform BIDS integrity check; S3. Preprocess and reconstruct the whole brain segmentation of T1-weighted sMRI data that have passed BIDS integrity check to obtain the cortical surface model and subcutaneous structure; S4, preprocessing the verified fMRI data, registering the preprocessed fMRI data to a cortical surface model reconstructed based on the T1-weighted sMRI data, and then converting the registered fMRI data and the T1-weighted sMRI data to MNI standard space; S5. Register the fMRI data in the MNI standard space to the Conte69 template, then downsample to the spatial resolution required by the CIFTI space. Then map the T1-weighted sMRI data in the MNI standard space and the downsampled fMRI data to the CIFTI grayscale space, thereby integrating the two in a unified coordinate system. Finally, encapsulate the T1-weighted sMRI and fMRI data in coordinate system 1 into a data format that meets the HCP standard.

2. The method for automatically partitioning a multimodal brain map of the Joint Human Connectome Project as claimed in claim 1, characterized in that: In the above-mentioned S1, magnetic field distribution information may be further obtained to correct image distortion caused by magnetic field inhomogeneity when pre-processing the verified fMRI data.

3. The method for automatically partitioning a multimodal brain map of the Joint Human Connectome Project as claimed in claim 1, characterized in that: In S2, BIDS Validator is used to perform BIDS integrity check on the T1-weighted sMRI data and fMRI data in NIFTI format. Only after the check passes can the subsequent steps be executed.

4. The method for automatically partitioning a multimodal brain map of the Joint Human Connectome Project as claimed in claim 1, wherein: In S3, the T1-weighted sMRI data are preprocessed and the whole brain segmentation is reconstructed using commands integrated in the FreeSurfer tool.

5. The method for automatically partitioning a multimodal brain map of the Joint Human Connectome Project as claimed in claim 1, wherein: In S4, the fMRIprep framework is used to preprocess and align the fMRI data, wherein the preprocessing operations include fMRI sequence slice calibration, rigid body calibration, and magnetic susceptibility distortion correction, and the registration algorithm is the bbregister algorithm; at the same time, in S4, the ANTs software is used to convert the aligned fMRI data and T1-weighted sMRI data into the MNI standard space.

6. The method for automatically partitioning a multimodal brain map of the Joint Human Connectome Project as claimed in claim 1, characterized in that: In S5, the CIFTIFY (Connectivity Informatics Technology Initiative Format Analysis) technology is used to encapsulate the T1-weighted sMRI and fMRI data into a data format that complies with the HCP standard.

7. An automatic partitioning system for multimodal brain maps in conjunction with the Human Connectome Project, characterized by: include: The data acquisition module is used to obtain T1-weighted sMRI data and fMRI data collected from the brain of the target subject, and the file format of both modal data is DICOM format; The conversion and verification module is used to convert T1-weighted sMRI data and fMRI data from DICOM format to NIFTI format that complies with the BIDS standard and perform BIDS integrity verification; The structural image processing module is used to preprocess and reconstruct the whole brain segmentation of T1-weighted sMRI data that have passed the BIDS integrity check to obtain the cortical surface model and subcutaneous structure; a functional image processing module, configured to preprocess the verified fMRI data, register the preprocessed fMRI data to a cortical surface model reconstructed based on the T1-weighted sMRI data, and then convert the registered fMRI data and T1-weighted sMRI data to the MNI standard space; The registration and encapsulation module is used to register the fMRI data in the MNI standard space to the Conte69 template, then downsample it to the spatial resolution required by the CIFTI space, and then map the T1-weighted sMRI data in the MNI standard space and the downsampled fMRI data to the CIFTI grayscale space, thereby integrating the two in a unified coordinate system. Finally, the T1-weighted sMRI and fMRI data in coordinate system 1 are encapsulated into a data format that conforms to the HCP standard.

8. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instruction is executed by a processor, it can implement the automatic partitioning method of the joint human connection group project multimodal brain map as described in any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that The storage medium stores a computer program, which, when executed by a processor, implements the automatic partitioning method of the joint human connectome project multimodal brain map as described in any one of claims 1 to 7.

10. A computer electronic device, characterized in that: including memory and processor; The memory is used to store computer programs; The processor is configured to implement the automatic partitioning method of the Joint Human Connectome Project multimodal brain map as described in any one of claims 1 to 7 when executing the computer program.

Images