Awesome AI Agents › Medical & Healthcare
MedRAX is an advanced AI-driven medical reasoning agent designed specifically for the interpretation of chest X-rays (CXRs). It integrates multiple state-of-the-art CXR analysis tools and multimodal large language models into a single unified framework, enabling it to address complex medical queries without the need for additional training. This versatility makes MedRAX a significant advancement in automated medical imaging analysis, particularly in clinical settings where comprehensive and accurate interpretation of CXRs is critical for disease management and patient care. The core architecture of MedRAX is built on the LangChain and LangGraph frameworks, utilizing GPT-4o with vision capabilities as its backbone large language model. It supports both local and cloud-based deployments and features a production-ready interface developed with Gradio. Its modular, tool-agnostic design allows for easy integration of new capabilities and tools, enhancing its adaptability and extensibility. MedRAX incorporates a variety of specialized tools for different aspects of CXR analysis: visual question answering (CheXagent and LLaVA-Med), segmentation (MedSAM and PSPNet), grounding (Maira-2), report generation (SwinV2 Transformer), disease classification (DenseNet-121), and synthetic X-ray generation (RoentGen). It also includes utilities for DICOM processing and visualization, making it a comprehensive solution for medical imaging workflows. To evaluate its performance, the developers introduced ChestAgentBench, a benchmark dataset comprising 2,500 complex medical queries across seven categories such as detection, classification, localization, comparison, relationship, diagnosis, and characterization. This benchmark is based on 675 expert-curated clinical cases and is designed to rigorously test multi-step reasoning capabilities in CXR interpretation. MedRAX is experimentally released with ongoing updates planned, including enhanced vision support for GPT-4o and MedSAM. It requires Python 3.8+, CUDA/GPU for optimal performance, and offers flexible tool selection to accommodate different resource constraints. Overall, MedRAX represents a cutting-edge, integrative approach to automated chest X-ray analysis, aiming to improve diagnostic accuracy and clinical decision-making.
https://github.com/bowang-lab/MedRAX
Multi-Agent Medical Assistant is an AI-powered chatbot system that integrates multi-agent orchestration, advanced retrieval-augmented generation, and medical image analysis to assist healthcare professionals, researchers, and patients with medical diagnosis and healthcare research.
TxAgent is an AI agent that leverages multi-step reasoning and a comprehensive toolbox of biomedical tools to provide personalized and precise therapeutic treatment recommendations based on real-time biomedical knowledge and clinical guidelines.
MedResearcher-R1 is a medical deep research agent plus the knowledge-informed trajectory synthesis pipeline that trained it, turning a domain knowledge graph into multi-hop question-answer pairs and tool-using reasoning trajectories.
MedAgents is a Multi-disciplinary Collaboration framework leveraging large language models to facilitate zero-shot medical reasoning through expert gathering, iterative analysis, and consensus-driven decision making.
A curated repository of research papers and resources reviewing the transformative applications and challenges of Large Language Models in medicine and healthcare.