Improving Hippocampus Volumetry Using a Deep Learning-based Proton Density-Weighted TSE sequence at 3T

Motivation: Accurate assessment of hippocampus volume is essential for early diagnosis, disease monitoring, understanding brain function, and guiding treatment decisions in a variety of neurological diseases, psychiatric conditions, and aging. Goal(s): To assess conventional and proposed high-resolution deep learning-based sequences in order to enable an accurate segmentation of the hippocampal subfields using 3T MRI. Approach: Utilization of a deep learning based high-resolution 2D proton density-weighted (hrPDW), conventional 3D T1-weighted and T2-weighted sequences for segmenting the hippocampus in Freesurfer. Results: As compared to conventional sequences, the proposed hrPDW improves the hippocampal contrast and provides accurate hippocampal segmentation. Impact: This study demonstrates that the deep learning based 2D high-resolution proton density weighted TSE sequence has a potential to reduce inaccuracies in hippocampus volumetry, which will ensure reliable diagnosis and monitoring of neurological and psychiatric conditions.