Methods, apparatus, devices, storage media, and program products for generating head models
By performing tissue segmentation and topology repair on MRI and MEG helmet data, a head model with high anatomical accuracy is generated, which solves the problems of low anatomical accuracy and non-robust registration in the existing head model construction, and realizes efficient and accurate head model generation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- GUOCI CLOUD DIGITAL (DEQING) TECHNOLOGY CO LTD
- Filing Date
- 2026-02-03
- Publication Date
- 2026-05-26
AI Technical Summary
Existing technologies for constructing head models in magnetoencephalography (MEG) and high-density electroencephalography (EEG) suffer from low anatomical accuracy, topological incompleteness of segmentation results, and non-robustness of the registration process. These issues lead to distorted or non-convergent computational results, low processing efficiency, and difficulty in deployment in diverse clinical and research environments.
By receiving MRI scans and MEG helmet data, tissue segmentation and topology repair are performed to generate a head model with high anatomical accuracy. Differential segmentation is performed using 3D U-Net and Otsu adaptive thresholding algorithms. Topologically complete masks are generated by combining morphological operations and Boolean operations. Registration and fusion are performed through spatial transformation matrices to improve automation and robustness.
It achieves efficient and accurate full-head tissue segmentation and registration, generating a head model that conforms to the anatomical structure, improving processing efficiency and geometric accuracy of the model, and solving the problems of segmentation accuracy and registration reliability in existing technologies.
Smart Images

Figure CN121661294B_ABST