CRL-2025: Multi-Class Segmentation Models for Human Fetal Brain MRI

Abdolmahdi Bagheri and the CRL2025 Fetal Atlas team developed these multi-class segmentation models for human fetal brain MRI analysis, updated in March 2026. The collection provides the trained weights and reference tools associated with the 'An MRI Atlas of the Human Fetal Brain' study to facilitate automated brain structure identification.

Use Cases

• Automated segmentation of fetal brain regions using the provided multi-class models
• Benchmarking fetal brain MRI analysis tools against the CRL-2025 reference atlas
• Integrating pre-trained segmentation weights into neuroimaging pipelines for prenatal diagnostics

Strengths

• Directly linked to the CRL-2025 fetal brain MRI atlas study
• Supports multi-class segmentation of human fetal brain structures
• Developed by specialized researchers at the CRL2025 Fetal Atlas organization

Limitations

• Requires specific citation of multiple academic papers for use
• Internal model architecture details depend on external GitHub documentation
• Restricted to fetal brain anatomy, limiting generalizability to adult neuroimaging

Provenance

Source

CRL2025 Fetal Atlas / arXiv preprint arXiv:2508.15034

Collection Method

Annotated MRI scans and model training

Freshness

Last updated March 2026.