Nonhuman Primate Brain Tissue Segmentation Using a Transfer Learning Approach

2025-03-28

AI Summary

Accurate segmentation of subcortical nuclei (e.g., thalamus, putamen) in non-human primate (NHP) brain MRI is hindered by scarce annotated data, low spatial resolution, and substantial interspecies anatomical variability. Method: We propose the first few-shot cross-species transfer segmentation framework tailored for NHP brain MRI—adapting the human-pretrained STU-Net model to NHP data. Our approach innovatively integrates STU-Net’s multi-scale contextual modeling with a dedicated cross-species feature alignment strategy, specifically designed for small-volume, low signal-to-noise-ratio NHP scans and interspecies structural heterogeneity. Contribution/Results: Under limited annotation budgets, our method achieves mean Dice > 0.88, IoU > 0.80, and HD95 < 7 mm, significantly improving robustness and generalizability for fine-grained brain region segmentation. This enables high-fidelity NHP brain atlasing, advancing mechanistic studies of neuropsychiatric disorders and cross-species mapping of human brain function.

Technology Category

• Humans and AI: Brain-Sensing and Analysis
• Computer Vision: Segmentation
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Web Mining and Content Analysis: Large pretrained models with web data
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Non-human primates (NHPs) serve as critical models for understanding human brain function and neurological disorders due to their close evolutionary relationship with humans. Accurate brain tissue segmentation in NHPs is critical for understanding neurological disorders, but challenging due to the scarcity of annotated NHP brain MRI datasets, the small size of the NHP brain, the limited resolution of available imaging data and the anatomical differences between human and NHP brains. To address these challenges, we propose a novel approach utilizing STU-Net with transfer learning to leverage knowledge transferred from human brain MRI data to enhance segmen-tation accuracy in the NHP brain MRI, particularly when training data is limited.The combination of STU-Net and transfer learning effectively delineates complex tissue boundaries and captures fine anatomical details specific to NHP brains. Notably, our method demonstrated improvement in segmenting small subcortical structures such as putamen and thalamus that are challenging to resolve with limited spatial resolution and tissue contrast, and achieved DSC of over 0.88, IoU over 0.8 and HD95 under 7. This study introduces a robust method for multi-class brain tissue segmentation in NHPs, potentially accelerating research in evolutionary neuroscience and preclinical studies of neurological disorders relevant to human health.

Problem

• Accurate NHP brain tissue segmentation despite limited annotated MRI data
• Overcoming anatomical differences between human and NHP brain MRI scans
• Improving segmentation of small subcortical structures with low resolution

Innovation

• STU-Net with transfer learning
• Leverages human MRI data
• Improves small structure segmentation