We explore the use of the Focal Tversky Loss function

We utilize transfer learning on the ADNI dataset, reduced in sample size by entropy selection to acheive state-of-the-art performance in multi-task classification. We provide a detailed analysis using class activation maps to demonstrate the model’s performance on neuropathological regions that are task-relevant and can help healthcare practicioners in interpreting the model’s decision.

A generalized focal loss function based on the Tversky index to address the issue of data imbalance in medical image segmentation.

A case study on factors that limit peformance of machine learning on biomedical images.