A two-step concept-based approach for enhanced interpretability and trust in skin lesion diagnosis
The main challenges hindering the adoption of deep learning-based systems in clinical settings are the scarcity of annotated data and the lack of interpretability and trust in these systems. Concept Bottleneck Models (CBMs) offer inherent interpretability by constraining the final disease prediction...
Збережено в:
| Автори: | , , |
|---|---|
| Формат: | Artigo |
| Мова: | Inglês |
| Опубліковано: |
Elsevier
2025-01-01
|
| Серія: | Computational and Structural Biotechnology Journal |
| Предмети: | |
| Онлайн доступ: | http://www.sciencedirect.com/science/article/pii/S2001037025000418 |
| Теги: |
Немає тегів, Будьте першим, хто поставить тег для цього запису!
|