PathoSage: New AI framework reduces pathology hallucinations by 40%

Pathology AI using Multimodal Large Language Models (MLLMs) often hallucinates morphological features, while agentic systems merge tool outputs and retrieved knowledge into shared contexts, making decisions vulnerable to conflicting evidence and context contamination. PathoSage directly addresses this with a three-stage framework that explicitly separates knowledge retrieval, evidence collection, and evidence adjudication for patch-level pathology multimodal reasoning. The core innovation, Structured Evidence Deliberation, independently evaluates heterogeneous evidence from tools, performs conflict analysis, and generates the final judgment in a fresh context to reduce anchoring bias.

Additionally, PathoSage introduces a training-free Beta-Bernoulli experience system with continuous credit assignment to model long-term tool reliability and construct similarity-weighted priors for future tool use. Experiments show it effectively mitigates VQA hallucinations and classifier disagreement, outperforming strong pathology MLLM and agentic baselines. The results highlight explicit evidence adjudication and reliability-aware tool modeling as key ingredients for robust pathology agents, potentially improving diagnostic accuracy in computational pathology.