Patho-AgenticRAG: Towards Multimodal Agentic Retrieval-Augmented Generation for Pathology VLMs via Reinforcement Learning

TL;DR

Patho-AgenticRAG introduces a multimodal retrieval-augmented generation framework for pathology VLMs, leveraging a database of pathology textbook embeddings to improve diagnostic accuracy and reduce hallucinations through joint text-image retrieval and reasoning.

Contribution

It presents a novel multimodal RAG system with a pathology textbook database enabling joint text-image search and reasoning, enhancing diagnostic performance in pathology VLMs.

Findings

Significantly outperforms existing models in pathology diagnosis tasks.

Supports joint text-image retrieval for better visual evidence utilization.

Improves accuracy in complex diagnostic scenarios.

Abstract

Although Vision Language Models (VLMs) have shown strong generalization in medical imaging, pathology presents unique challenges due to ultra-high resolution, complex tissue structures, and nuanced clinical semantics. These factors make pathology VLMs prone to hallucinations, i.e., generating outputs inconsistent with visual evidence, which undermines clinical trust. Existing RAG approaches in this domain largely depend on text-based knowledge bases, limiting their ability to leverage diagnostic visual cues. To address this, we propose Patho-AgenticRAG, a multimodal RAG framework with a database built on page-level embeddings from authoritative pathology textbooks. Unlike traditional text-only retrieval systems, it supports joint text-image search, enabling direct retrieval of textbook pages that contain both the queried text and relevant visual cues, thus avoiding the loss of critical image-based information. Patho-AgenticRAG also supports reasoning, task decomposition, and multi-turn search interactions, improving accuracy in complex diagnostic scenarios. Experiments show that Patho-AgenticRAG significantly outperforms existing multimodal models in complex pathology tasks like multiple-choice diagnosis and visual question answering. Our project is available at the Patho-AgenticRAG repository: https://github.com/Wenchuan-Zhang/Patho-AgenticRAG.