Keywords: Fetal, Brain, Data Analysis, Surface Reconstruction
Motivation: Fetal cortex undergoes rapid development. Deep learning models are crucial tools for cortical surface reconstruction and morphometry. However, these models have shown limited performance due to the scarcity of fetal brain surface data and the fast-paced changes during development.
Goal(s): To develop a high-performance fetal cortical surface reconstruction model.
Approach: A segmentation nn-Unet model generated trustworthy brain tissue labels to create inner and outer surfaces, which were then used to train the proposed multi-input KAN model, CortexKAN, using an ODE-based training strategy.
Results: CortexKAN outperforms existing models in fetal cortical surface reconstruction.
Impact: CortexKAN model achieves superior performance in fetal cortical surface reconstruction, outperforming existing methods. Additionally, a training strategy without cortical labels was demonstrated effective, enabling accurate reconstruction when surface labels are unavailable.
How to access this content:
For one year after publication, abstracts and videos are only open to registrants of this annual meeting. Registrants should use their existing login information. Non-registrant access can be purchased via the ISMRM E-Library.
After one year, current ISMRM & ISMRT members get free access to both the abstracts and videos. Non-members and non-registrants must purchase access via the ISMRM E-Library.
After two years, the meeting proceedings (abstracts) are opened to the public and require no login information. Videos remain behind password for access by members, registrants and E-Library customers.
Keywords