MedExpMem: Adapting Experience Memory for Differential Diagnosis

Researchers have developed MedExpMem, an experience memory framework that enables medical vision-language models to accumulate differential diagnosis expertise from their own diagnostic failures. The system, which mirrors physician learning by identifying knowledge gaps and refining understanding through reflective re-diagnosis, improved diagnostic accuracy by up to 7.0% across 11 radiology subspecialties. This approach addresses a key limitation of current medical AI models, which store static knowledge that does not evolve across diagnostic encounters.

arXiv:2605.22872v1 Announce Type: new Abstract: Experienced physicians develop diagnostic expertise through clinical practice, acquiring not only disease knowledge but also the ability to differentiate confusable conditions. Current medical vision-language models VLMs lack this capability -- their parameters encode static knowledge that does not evolve across diagnostic encounters. We propose MedExpMem, an experience memory framework enabling VLM-based diagnostic agents to accumulate differential diagnosis expertise. Unlike retrieval-augmented generation, which retrieves encyclopedic disease descriptions, MedExpMem memorizes discriminative experience derived from the agent's own diagnostic failures and organizes them as pairwise differential notes encoding key discriminators, actionable decision rules and reasoning error patterns. The framework adopts a two-phase construction process mirroring physician learning: initial practice exposes knowledge gaps, and reflective re-diagnosis refines understanding. When encountering new cases, the agent retrieves experience memory to guide differential reasoning. We evaluate MedExpMem on a radiology benchmark spanning 11 subspecialties. Results demonstrate consistent accuracy improvements, maximum 7.0%, across diverse models and scales. Analytical experiments validate experience quality and robustness, demonstrating MedExpMem as a competitive method addresses medical adaptation needs beyond the reach of parameteric learning.