CARE-MVLM: Counterfactual Abstention and Region-Grounded Evidence in Mammography Vision-Language Models
Qu, B.; Liu, W.; Murrow, M.; Burger, M.; Guo, X.; Vaidya, M. S.; Rose, S. L.; Kantarcioglu, M.; Malin, B. A.; Yin, Z.
Show abstract
Mammography interpretation requires tying each finding to supporting evidence and withholding judgment when that evidence is unavailable. Vision-language models are increasingly used for image-based medical question answering, yet existing benchmarks suggest that they struggle to jointly support accurate answers, evidence grounding, and appropriate abstention. We propose CARE-MVLM, built on Qwen2.5-VL-7B-Instruct, to address this gap through separate modules for answer prediction, answer-conditioned evidence generation, and answerability selection. On 7,358 triplets curated from CBIS-DDSM, MIAS, and VinDr-Mammo, CARE-MVLM substantially improves lesion-sensitive selective behavior over same VLM backbone baselines and exhibits markedly stronger lesion-sensitive abstention than GPT-5.6 and Gemini-3.5-flash while attaining the highest joint answer, grounding, and abstention reliability in the comparison.
Matching journals
The top 4 journals account for 50% of the predicted probability mass.
Similar papers in this journal
Similar papers in this journal
Similar papers in this journal
- Generative AI Enables Medical Image Segmentation in Ultra Low-Data Regimes 94%
- Features fusion or not: harnessing multiple pathological foundation models using Meta-Encoder for downstream tasks fine-tuning 94%
- Accurate recognition of colorectal cancer with semi-supervised deep learning on pathological images 93%
Similar papers in this journal
- Deep learning inference of cell type-specific gene expression from breast tumor histopathology 91%
- Generalizing AI-driven Assessment of Immunohistochemistry across Immunostains and Cancer Types: A Universal Immunohistochemistry Analyzer 91%
- Real-World Benchmarking and Validation of Foundation Model Transformers for Endometrial Cancer Subtyping from Histopathology 90%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.