MIT and collaborators developed a technique to make computer vision models, including those used in medical imaging, provide clearer, concept-based explanations for their predictions.
Key Details
- 1The new method extracts and uses concepts learned by the model during training for explanations, instead of relying solely on human-defined concepts.
- 2A sparse autoencoder and multimodal large language model are used to extract and describe these concepts in plain language.
- 3The approach limits the model to using five concepts per prediction for clarity and relevance.
- 4Compared to state-of-the-art concept bottleneck models (CBMs), this technique achieved higher accuracy and more precise explanations in tasks including medical image diagnosis.
- 5The research will be presented at the International Conference on Learning Representations, with future plans to scale the approach and address information leakage.
Why It Matters
Improved explainability is critical for the adoption and trust of AI models in medical diagnostics, where human-understandable reasoning can support clinicians' decision-making. Techniques that balance accuracy with transparency help meet regulatory and ethical demands in radiology AI.
Source
EurekAlert
Related News
•EurekAlert
AI Pathology Tool SÉMIL Improves Stage II Bowel Cancer Risk Assessment
A La Trobe University-developed AI tool accurately predicts relapse risk in stage II bowel cancer using digital pathology images and descriptions.
•EurekAlert
AI Tool Predicts Which Rectal Cancer Patients Benefit from Intensive Therapy
UCL researchers developed an AI that analyzes biopsy slides to identify rectal cancer patients who benefit from adding irinotecan to standard chemoradiotherapy.
•EurekAlert
AI-Guided Handheld Cardiac Ultrasound Reduces Referrals and Costs in Spain
AI-guided handheld cardiac ultrasound enables primary care physicians to detect heart failure, reducing specialist referrals and saving costs.