Publications by authors named "Tomoki Hakotani"
Article Synopsis
- Deep learning techniques have advanced the automated and precise analysis of musculoskeletal (MSK) structures in medical images, though many current methods are limited to 2D images and small datasets.
- This study focused on validating an enhanced deep learning model for volumetric segmentation of hip and thigh structures in CT images, utilizing a diverse set of databases to ensure broader applicability.
- The improved model demonstrated superior segmentation accuracy and reliable failure detection through predictive uncertainty, indicating its potential for effective analysis of MSK structures in large-scale clinical settings.
