ProtoMedAgent: Multimodal Clinical Interpretability via Privacy-Aware Agentic Workflows
Quick Answer
ProtoMedAgent introduces a novel framework for multimodal clinical reporting, achieving 91.2% Comparison Set Faithfulness in a 4,160-patient cohort, significantly outperforming standard RAG's 46.2%.
Quick Take
It employs a neuro-symbolic approach with a privacy gate to ensure data security and reduce inference risks by 9.8%.
Key Points
- ProtoMedAgent formalizes clinical reporting as a zero-gradient optimization problem.
- It distills visual and tabular features into a discrete semantic memory.
- Achieves 91.2% faithfulness compared to standard 's 46.2% performance.
- Introduces a semantic privacy gate governed by k-anonymity and l-diversity.
- Reduces artifact-level membership inference risks by 9.8%.
Paper Resources
Article Content
From source RSS / original summaryarXiv:2605. 14113v1 Announce Type: new Abstract: While interpretable prototype networks offer compelling case-based reasoning for clinical diagnostics, their raw continuous outputs lack the semantic structure required for medical documentation. Bridging this gap via standard (RAG) routinely triggers ``retrieval sycophancy,'' where (LLMs) hallucinate post-hoc rationalizations to align with visual predictions.
We introduce ProtoMedAgent, a framework that formalizes multimodal clinical reporting as an iterative, zero-gradient test-time optimization problem over a strict neuro-symbolic bottleneck. Operating on a frozen prototype backbone, we distill latent visual and tabular features into a discrete semantic memory. Online generation is strictly constrained by exact set-theoretic differentials and a reflective Scribe-Critic loop, mathematically precluding unsupported narrative claims.
To safely bound data disclosure, we introduce a semantic privacy gate governed by $k$-anonymity and $\ell$-diversity. Evaluated on a 4,160-patient clinical cohort, ProtoMedAgent achieves 91. 2\% Comparison Set Faithfulness where it fundamentally outperforms standard RAG (46. 2\%). ProtoMedAgent additionally leverages a binding $\ell$-diversity phase transition to systematically reduce artifact-level membership inference risks by an absolute 9. 8\%.
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CV
See more →ProMoE-FL: Prototype-conditioned Mixture of Experts for Multimodal Federated Learning with Missing Modalities
ProMoE-FL introduces a Prototype-conditioned Mixture-of-Experts framework for multimodal federated learning, effectively addressing missing modalities. It outperforms existing methods on four chest X-ray datasets, demonstrating superior feature synthesis capabilities in both homogeneous and heterogeneous settings.