Life & medical sciences · OECD FoS 3.2 · China MoE 1002 临床医学
Clinical medicine
临床医学
L2Local breakthroughs ↑ Rising Score 2.15 · Last reviewed 2026-10-21 · Next review 2027-01-19
BIGGEST BREAKTHROUGH 2025-04-09 AI · majorField · major
AMIE conversational diagnostic AI (Nature) ↗
Beats PCPs on diagnostic accuracy in simulated consults
Why this level
FDA lists 1,614 AI-enabled devices (4 Sep 2026 update); AMIE out-diagnosed clinicians in simulated consults (Nature 2025). Prospective evidence for LLMs in care is still thin.
S = 0.30·method + 0.35·discovery + 0.25·adoption + 0.10·education (education scores 0 when there is no evidence; if empty, the other weights are renormalised)
Sources
- fda.gov/medical-devices/digital-health-center-excellence/artificial-intelligence-enabled-medical-devices
- nature.com/articles/s41586-025-08866-7
Composite 1.5–2.49: at least one peer-reviewed/officially verified AI-key result, or one dimension at 3; routine workflow largely unchanged. All levels →
Research directions
Diagnostic dialogue agentsClinical documentationRisk predictionMedical LLM evaluation
Papers
- Human–artificial intelligence collaboration for comanaging epilepsy in real time
AI: an LLM talks with patients about EEG biomarkers and retrains its detector from feedback; Field: devices move from passive recording to real-time co-management (inpatient prototype)
2026-10-13 · Nature Medicine
- Generative AI-enabled clinical decision support system in primary care: a pragmatic, cluster-randomized trial
9,691-patient trial: safe but no significant benefit
2026-06-26 · Nature Medicine
- Automation Bias in Large Language Model–Assisted Diagnostic Reasoning among Physicians Trained in AI Literacy — A Randomized Clinical Trial
Shows risks in human–AI diagnostic teaming
2026-04-23 · NEJM AI
- Towards conversational diagnostic artificial intelligence
Beats PCPs on diagnostic accuracy in simulated consults
2025-04-09 · Nature
- Randomized Trial of a Generative AI Chatbot for Mental Health Treatment
Significant symptom reductions vs waitlist in 210 adults
2025-03-27 · NEJM AI
- Large language models encode clinical knowledge
First LLM to pass USMLE-style questions
2023-07-12 · Nature
- High-performance medicine: the convergence of human and artificial intelligence
Influential roadmap for AI in clinical medicine
2019-01-07 · Nature Medicine
- Pivotal trial of an autonomous AI-based diagnostic system for detection of diabetic retinopathy in primary care offices
Evidence behind the first autonomous AI diagnostic clearance
2018-08-28 · npj Digital Medicine
- Overview of CHIP 2026 Shared Task 1: Patient-Specific Phenotype Concept Recognition in Full-Text Biomedical Literature
Overview of a shared task on recognising phenotype concepts and linking them to specific patients in full-text biomedical papers.
2026-10-21 · arXiv · Xuanlyu Zhu, Feng Tian, Yang Zhong, Pengqi Zheng
- Reading the Mask: Descriptor-Grounded BI-RADS Rationale Generation for Breast MRI Segmentation
Derives BI-RADS descriptors from breast DCE-MRI segmentation masks to generate radiologist-style rationales.
2026-10-21 · arXiv · Hrishikesh Singh, Aarti Sharma, Divyanshu Kumar, Millie Pant
- Towards Early Warning Systems for Shigellosis in Africa: Multi-Horizon Forecasting of Incidence in Rural Western Kenya Using Statistical State-Space and Machine Learning Models
Compares state-space and ML models for multi-horizon forecasting of childhood shigellosis in rural western Kenya using 16 years of data.
2026-10-21 · medRxiv · Ogwel, B., Onyando, B. O., Nyawanda, B. O., Awuor, A. O.
- From Omics Data to Biological Context: The CURIE AI En-gine. Illustrated with the Transcriptomic Response to Bacte-rial Pneumonia
Demonstrates an AI engine that contextualises transcriptomic results, using blood RNA-seq from bacterial pneumonia patients.
2026-10-21 · medRxiv · Castellani, G., Rodriguez-Mena, A., Tarrago-Claramunt, X., Madarnas-Gomez, V.
- When to Stop Asking: Factor Coverage Bias and LLM Determinability Judgments with Underspecified Information in Medicine
Finds LLMs judge clinical determinability by amount of information rather than whether missing facts could change the decision.
2026-10-19 · arXiv · Yusuke Watanabe, Yohei Kobashi, Taisei Tosaki, Takeshi Kojima
- Deep learning-based electrocardiographic differentiation of Takotsubo syndrome from acute anterior myocardial infarction: explainability, human-AI interaction, and a utilization bias hypothesis
Deep-learning ECG model distinguishes Takotsubo from anterior STEMI; study measures its effect on readers of different backgrounds.
2026-10-19 · medRxiv · Kagawa, E., Kato, M., Hidaka, T., Kunita, E.
- From meal text to carbohydrate to postprandial glucose rise: QLoRA fine-tuning of the Qwen3-4B language model and gradient-boosting prediction in type 2 diabetes
Tests whether a fine-tuned small open LLM estimates meal carbohydrates from text and improves post-meal glucose prediction in T2D.
2026-10-19 · medRxiv · Bou Ali, M. N., Lassoued, N., Ben Amor, B., Nasr, S.
- Deep Learning-Based Anatomical Level Classification of Craniofacial CT Slices: Comparison of CNN Architectures and Ensemble Methods
Classifies craniofacial CT slices into six anatomical levels using 11,769 images from 10 Japanese university hospitals.
2026-10-19 · medRxiv · Nishimoto, S., Nakajima, K., Yoshiike, H., Ishise, H.
- From arXiv, bioRxiv, medRxiv and OpenAlex (last ~30 days); summaries written by AIMap from the abstract.
- X-Ragent: Tool-Augmented Reinforcement Learning for Evidence-Grounded Chest X-Ray Reasoning 2026-10-12
- Scoring Higher, Answering Worse: Mitigating Reward Hacking in Rubric-Based RL via Protocol-Level Rubrics 2026-09-30
- On-Policy or Off-Policy Learning? A Systematic Study of Distillation Dynamics 2026-09-28
- Where Does Retrieval-Based Open-Ended Evaluation Fail? Automatic Taxonomy Induction from Long-Form Medical Answer Factuality Verification 2026-09-24
- From Pattern Recognizers to Personalized Companions: A Survey of Large Language Models in Mental Health 2026-09-21
- All related papers in the archive →
Milestones 10
Big glowing nodes are leaps: AI leap cool · Field leap warm · double ring = both
-
Clinical
Human–AI collaboration for comanaging epilepsy in real time (Nature Medicine)AI: an LLM talks with patients about EEG biomarkers and retrains its detector from feedback; Field: devices move from passive recording to real-time co-management (inpatient prototype)
Nature Medicine -
Clinical
Kenya cluster-RCT: LLM decision support did not reduce 14-day treatment failure (Nature Medicine)9,691-patient trial: safe but no significant benefit
Nature Medicine -
Clinical
Automation bias in LLM-assisted diagnosis among AI-literate physicians (NEJM AI RCT)Shows risks in human–AI diagnostic teaming
NEJM AI -
Paper
AMIE conversational diagnostic AI (Nature)Beats PCPs on diagnostic accuracy in simulated consults
Nature -
Clinical
Therabot: first RCT of a generative-AI therapy chatbot (NEJM AI)Significant symptom reductions vs waitlist in 210 adults
NEJM AI -
Paper
Med-PaLM: LLMs encode clinical knowledge (Nature)First LLM to pass USMLE-style questions
Nature -
Paper
Topol: High-performance medicine, the convergence of human and AI (Nature Medicine)Influential roadmap for AI in clinical medicine
Nature Medicine -
Clinical
Pivotal trial of autonomous AI for diabetic retinopathy in primary care (npj Digit. Med.)Evidence behind the first autonomous AI diagnostic clearance
npj Digital Medicine -
Regulatory
FDA authorises first autonomous AI diagnostic, IDx-DRFirst FDA-authorised autonomous AI diagnostic
FDA De Novo DEN180001
No milestones match this filter.
More papers & repos 8
Latest news 12
- AZ and Oracle team up on AI platform for clinical trialspharmaphorumen
- A Controlled LLM Workflow for Evidence-Grounded Diagnosis of Risk Score DifferencesarXiv cs.AIenPaperPreprint
- AMA Introduces Governance Framework for Agentic AI in Clinical CareDigital Health Newsen
- Diagnosis Is Beside the PointMedCity Newsen