[Nature Reviews Bioengineering🔥] Application of Large Language Models in Medicine. A curated list of practical guide resources of Medical LLMs (Medical LLMs Tree, Tables, and Papers)

ClinVec: Unified Embeddings of Clinical Codes Enable Knowledge-Grounded AI in Medicine

[ NeurIPS 2022 ] Official Codebase for "ETAB: A Benchmark Suite for Visual Representation Learning in Echocardiography"

OpenEvidence MCP: open-source browser-session MCP server for human and AI-agent medical workflows

AI-powered chest X-ray pneumonia detection with 86% accuracy and 96.4% sensitivity, validated on an independent (cross-operator) cohort of 485 pediatric samples. Built with TensorFlow & FastAPI.

Systematic evaluation of hallucination risks in Large Language Models (GPT-4, Claude 3, Gemini Pro) for clinical proteomics and mass spectrometry interpretation. Production-ready detection framework with comprehensive benchmarks.

Multi-provider LLaMA 3.1 70B benchmark for dental knowledge — latency, consistency, and factual accuracy across 7 hosts, tracked in Weights & Biases.

Ethnic bias analysis in medical imaging AI: Demonstrating that explainable-by-design models achieve 80% bias reduction across 5 ethnic groups (50k images)

🩺 AstraMed: A Clinical Risk Intelligence Platform. Powered by SOTA Ensemble ML & BioMistral-7B for predictive medical analytics and explainable risk scoring.

An application to monitor clinical AI models

Rare AI Archive: open-source agentic diagnostic AI for rare genetic diseases — decentralized post-training, clinician validation, federated deployment

MobileNetV2 pneumonia classifier validated on an independent 485-sample cross-operator cohort. 96.4% sensitivity, 96.4% ROC-AUC, bootstrap p=0.978. FastAPI inference API, Streamlit dashboard, DICOM support, Docker-ready

AI-assisted orthodontic treatment planning pipeline for clear aligners with dual-agent deep learning and 4D staging simulation.

An interpretable ML system for diabetes risk prediction using clinical data. Features SHAP explanations, model comparison (Random Forest vs XGBoost), and a deployment-ready pipeline. Achieves 0.85 AUC with clinical decision support.

Turn clinical guideline PDFs into deterministic, auditable decision engines with source citations.

Full-stack clinical AI platform — multi-agent RAG over FHIR patient data with 20+ medical tools, hybrid vector search, and real-time streaming

AI-powered ICU early warning system predicting sepsis, hypotension, and hemodynamic collapse 2-6 hours before onset. 4-engine ML ensemble (LightGBM, XGBoost, GRU-D, TCN) with physics-based safety net and multi-agent LLM clinical debate.

AI-powered thermal imaging system for early neonatal sepsis detection in NICUs. Non-invasive monitoring using deep learning to identify life-threatening infections before clinical symptoms appear. Collaboration with General University Hospital of Patras & Universitat Autònoma de Barcelona UAB.

Production-grade Medical Device IoT Platform. Real-time Clinical AI anomaly detection. FHIR R4 healthcare interoperability. Edge Gateway processing. Modbus → MQTT → FHIR integration. SignalR dashboards. ASP.NET Core innovation.

Comprehensive literature review on hybrid AI systems combining temporal knowledge graphs, clinical constraints, and generative models for emergency department decision support