Brain Region Boundary Recommendation System Based on Deep Learning

Mapping the brain is a crucial task in brain science research. However, manual brain mapping requires highly specialized skills and consumes significant time and resources to process large-scale data. This paper proposes an automated brain region boundary recommendation system to address the challenge. The system can automatically recommend subsequent monkey brain slices based on the baseline brain map provided by the researcher. The system utilizes deep metric learning to extract features from image patches. This approach can extract richer and more discriminative feature information compared to traditional methods, resulting in improved system recommendation accuracy. The recommendation results demonstrate the system's high accuracy in automated brain mapping and its ability to significantly reduce researchers' workload.