Adaptive Diagnostic Reasoning Framework for Pathology with Multimodal Large Language Models

2025-11-14

AI Summary

Current pathological AI tools predominantly operate as black-box models, lacking auditable and interpretable reasoning processes—hindering clinical adoption. To address this, we propose RECAP-PATH, a novel framework built upon multimodal large language models (MLLMs) that enables evidence-driven autonomous reasoning in pathological diagnosis for the first time, shifting AI from passive pattern recognition to traceable, verifiable diagnostic decision-making. We introduce an innovative two-stage self-learning paradigm that autonomously distills diagnostic criteria from minimal labeled data—without parameter updates—and jointly models visual perception and rationale generation. Evaluated on breast and prostate cancer datasets, RECAP-PATH achieves statistically significant improvements in diagnostic accuracy over state-of-the-art methods. Moreover, its generated diagnostic rationales exhibit strong agreement with expert assessments (Cohen’s κ > 0.85), substantially enhancing clinical trustworthiness and interpretability.

Technology Category

• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

AI tools in pathology have improved screening throughput, standardized quantification, and revealed prognostic patterns that inform treatment. However, adoption remains limited because most systems still lack the human-readable reasoning needed to audit decisions and prevent errors. We present RECAP-PATH, an interpretable framework that establishes a self-learning paradigm, shifting off-the-shelf multimodal large language models from passive pattern recognition to evidence-linked diagnostic reasoning. At its core is a two-phase learning process that autonomously derives diagnostic criteria: diversification expands pathology-style explanations, while optimization refines them for accuracy. This self-learning approach requires only small labeled sets and no white-box access or weight updates to generate cancer diagnoses. Evaluated on breast and prostate datasets, RECAP-PATH produced rationales aligned with expert assessment and delivered substantial gains in diagnostic accuracy over baselines. By uniting visual understanding with reasoning, RECAP-PATH provides clinically trustworthy AI and demonstrates a generalizable path toward evidence-linked interpretation.

Problem

• Developing interpretable AI framework for pathology diagnostics
• Shifting AI from pattern recognition to evidence-based reasoning
• Generating human-readable rationales for clinical decision auditing

Innovation

• Self-learning paradigm for diagnostic reasoning
• Two-phase learning autonomously derives diagnostic criteria
• Generates cancer diagnoses without weight updates