ACL2026

MDTeamGPT: Mitigating Context Collapse and Enabling Self-Evolution in Medical Multi-Agent Reasoning

Kai Chen, Xinfeng Li, Tianpei Yang, Hewei Wang, Guang Yang, Jing Huo, Yang Gao

Abstract

Large language models (LLMs) have shown great potential in multi-disciplinary team (MDT) medical consultations. However, long, multi-round, multi-role interaction trajectories inevitably lead to severe information dilution and context window overload, triggering context collapse which destabilizes reasoning. Furthermore, prior systems typically rely on unstructured trajectory history storage without structurally distilling key information or reflecting on errors, severely limiting continuous learning capabilities. We propose MDTeamGPT, a context-resilient and self-evolving multi-agent framework. Mechanistically, we introduce a specialized Lead Physician mechanism combined with a Residual Context architecture to compress and reorganize multi-round consensus, effectively mitigating context overload and reducing computational costs. For memory, we design a Dual Knowledge Base system comprising a Correc-tKB for verified trajectories and a ChainKB for reflective error analysis, enabling self-evolution via retrieval from both successes and failures. We evaluated our framework on standard text datasets (MedQA, PubMedQA), multimodal benchmarks (VQA-RAD, SLAKE), and collected more complex clinical problems. Experimental results show that MDTeamGPT substantially outperforms existing baselines across both text-based and multimodal tasks, while also demonstrating superior diagnostic performance and stability in complex clinical scenarios. Our code is available at GitHub repository. A 35-year-old man with IgA nephropathy presented with confusion, blurry vision, and seizures. Two weeks before presentation, he had started receiving cyclosporine. Physical examination was notable for a blood pressure of 160/80 mm Hg, drowsiness, and decreased visual acuity. A fundoscopic examination was normal. T2-weighted MRI with fluid-attenuated inversion recovery sequencing of the head was performed. What is the most likely diagnosis? A. Acute demyelinating encephalomyelitis B. Methanol ingestion C. Multifocal ischemic infarcts D. Posterior reversible encephalopathy syndrome E. West Nile virus encephalitis Arrange Doctors