NEJM AI
LightNews LightNews
banner
ai.nejm.org
NEJM AI
@ai.nejm.org
1.2K followers 91 following 290 posts
NEJM AI, a new monthly journal from the publisher of @nejm.org, explores the cutting-edge applications of artificial intelligence and machine learning in clinical medicine. Online at ai.nejm.org.
Posts Replies Media Videos
ai.nejm.org
When the tool outperforms the team: on NEJM AI Grand Rounds, Dr. Jonathan Chen discusses why human–AI collaboration must be tested, not assumed. GPT‑4 changed the scoreboard. Learn more: nejm.ai/ep35

#AI #MedSky #MLSky
November 11, 2025 at 2:33 PM
ai.nejm.org
“A Case Study of AI-Enabled Software as a Medical Device Cleared by the FDA for Assessing Hemorrhage Risk Index (APPRAISE-HRI) after Trauma” by A. Frock et al. nejm.ai/48yUv5p

#AI #MedSky #MLSky
Figure 1. APPRAISE-HRI Software as a Medical Device. Figure 2. Step-By-Step Exclusion Process of Trauma Patients from the Stanford Study, Where We Collected the Data between August 2020 and August 2021. Figure 3. Step-By-Step Exclusion Process of Trauma Patients from the LITES Study, Where We Collected the Data between January 2017 and June 2019.
November 10, 2025 at 1:31 PM
ai.nejm.org
A new Perspective reviews insights from the 2025 Symposium on AI for Learning Health Systems, examining how AI tools have moved from pilot studies to widespread use in clinical care. Learn more: nejm.ai/4hgF5F4

#AI #MedSky #MLSky
“Are humans the ideal examiners of LLMs’ behavior? Until a coherent regulatory pathway emerges, clinicians and health systems will shoulder the responsibility and liability for ensuring these tools benefit rather than harm patients.” Perspective “The Moment AI Arrived in the Clinic: Insights from the SAIL 2025 Year in Review” by P. Elias et al.
November 7, 2025 at 2:15 PM
ai.nejm.org
A new Perspective argues that the term “AI hallucination” is a misnomer, contending that large language model errors more closely resemble “confabulation” — active, fabricated outputs — than hallucination, which implies conscious perception. Learn more: nejm.ai/4nV6JKh

#AI #MedSky #MLSky
Figure 1. A Man Asleep Dreaming of Monsters. Aquatint by Francisco de Goya, 1796/98. Reprinted with permission from the Wellcome Collection.
November 6, 2025 at 1:30 PM
ai.nejm.org
“I do not trust machine learning.” On NEJM AI Grand Rounds, Dr. Jonathan Chen reminds clinicians that real care needs reliability, not vibes. Listen to the full episode: nejm.ai/ep35

#AI #MedSky #MLSky
November 5, 2025 at 2:10 PM
ai.nejm.org
A new Perspective summarizes the U.S. National Academy of Medicine’s AI Code of Conduct special publication, highlighting the critical need to establish shared guardrails to ensure the best health outcomes for the public. Learn more: nejm.ai/3Jf0MJ1

#AI #MedSky #MLSky
“As novel AI tools and technologies proliferate across health and health care, it is more critical than ever for organizations to prioritize safety, transparency, shared governance, and alignment with best practices in health AI.” Perspective “Advancing the National Academy of Medicine Artificial Intelligence Code of Conduct for Health and Medicine” by L.L. Adams et al.
November 4, 2025 at 2:31 PM
ai.nejm.org
A new Case Study evaluates APPRAISE-HRI, an AI-enabled clinical decision support tool trained on vital signs of trauma patients to stratify hemorrhage risk into three levels that was cleared by the U.S. Food and Drug Administration. Learn more: nejm.ai/48yUv5p

@fda.gov #AI #MedSky #MLSky
Figure 1. APPRAISE-HRI Software as a Medical Device.
November 3, 2025 at 2:20 PM
ai.nejm.org
Comfort or conviction? Dr. Jonathan Chen left a good job for purpose — then worked harder than ever. The payoff: no regrets.

Listen to the latest episode of the NEJM AI Grand Rounds podcast: nejm.ai/ep35

#AI #MedSky #MLSky
October 31, 2025 at 1:15 PM
ai.nejm.org
Datasets, Benchmarks, and Protocols by L.G. McCoy et al.: Assessment of Large Language Models in Clinical Reasoning: A Novel Benchmarking Study nejm.ai/3KlD5yX

@liamgmccoy.bsky.social #AI #MedSky #MLSky
Figure 2. Overall Performance of Models on Full Script Concordance Testing Benchmark. Figure 3. Per-Test Performance of Models and Humans of Varying Training Levels. Figure 4. Comparison of Optimal Response Distribution with Overall Model Response Distribution.
October 30, 2025 at 12:10 PM
ai.nejm.org
Case Study by E. Klang et al.: Assessing Retrieval-Augmented Large Language Models for Medical Coding nejm.ai/4mxIQH6

#AI #MedSky #MLSky
Figure 1. LLM Performance in ICD Code Generation from Clinical Notes. Figure 2. Comparative Evaluations of ICD-10-CM Code Assignments: RAG-Enhanced LLMs versus Human Coders.
October 29, 2025 at 1:15 PM
ai.nejm.org
Volume 2, No. 11 of NEJM AI is now available! Here is a preview of the latest content:

𝗘𝗱𝗶𝘁𝗼𝗿𝗶𝗮𝗹
Humanity’s Next Medical Exam: Preparing to Evaluate Superhuman Systems nejm.ai/43FVI7x

#MedSky #AI #MLSky
The cover of the November 2025 issue of NEJM AI with "NEW ISSUE NOW AVAILABLE" above it
October 28, 2025 at 5:43 PM
ai.nejm.org
Raffi Hagopian, MD, responds to a letter regarding the Original Article he coauthored, “AI Opportunistic Coronary Calcium Screening at Veterans Affairs Hospitals.” Read the full response: nejm.ai/46QxtWh

#AI #MedSky #MLSky
Page 1 of "A Response to Dr. Ro’s Letter Regarding the Original Article, 'AI Opportunistic Coronary Calcium Screening at Veterans Affairs Hospitals'" Bottom: Read full article at ai.nejm.org.
October 28, 2025 at 1:35 PM
ai.nejm.org
Policy Corner by David Blumenthal, MD, MPP: The Trump Administration’s Recent Policy Proposals Regarding Artificial Intelligence nejm.ai/46Qwqpj

#AI #MedSky #MLSky
Page 1 of the Policy Corner article "The Trump Administration’s Recent Policy Proposals Regarding Artificial Intelligence" Bottom: Read full article at ai.nejm.org.
October 27, 2025 at 12:31 PM
ai.nejm.org
Perspective by Daniel Restrepo, MD: Raging against the Machine: The Human Side of the Story nejm.ai/46W1Fzi

#AI #MedSky #MLSky
Page 1 of the Perspective "Raging against the Machine: The Human Side of the Story" Bottom: Read full article at ai.nejm.org.
October 24, 2025 at 4:15 PM
ai.nejm.org
A new study introduces a publicly available script concordance test benchmark that reveals important limitations in the clinical reasoning capabilities of leading LLMs, particularly in processing new information under uncertainty. nejm.ai/3KlD5yX

@liamgmccoy.bsky.social #AI #MedSky #MLSky
Figure 3. Per-Test Performance of Models and Humans of Varying Training Levels.
October 24, 2025 at 1:15 PM
ai.nejm.org
A new study explores AI models with retrieval-augmented generation and suggests the potential for reducing administrative burden in health care, increasing coding efficiency, and improving clinical documentation. Full study results: nejm.ai/4mxIQH6

#AI #MedSky #MLSky
Figure 1. LLM Performance in ICD Code Generation from Clinical Notes.
October 23, 2025 at 12:31 PM
ai.nejm.org
“Passion” isn’t a prerequisite. On NEJM AI Grand Rounds, Dr. Jonathan Chen describes growing into medicine — and why honesty about motivation helps real patients, not résumés. Hear more from Dr. Chen in the full episode: nejm.ai/ep35

#AI #MedSky #MLSky
October 22, 2025 at 4:15 PM
ai.nejm.org
A new article by David Blumenthal, MD, MPP, analyzes the Trump administration’s executive orders and accompanying action plan on AI, highlighting their potential effects on health care research, innovation, and clinical practice in the U.S. and globally. nejm.ai/46Qwqpj

#AI #MedSky #MLSky
“A ‘try first’ orientation seems to run directly counter to medicine’s long-accepted ethical imperative to ‘first, do no harm,’ and may leave many health professionals uncomfortable with rapidly deployed AI applications.” Policy Corner “The Trump Administration’s Recent Policy Proposals Regarding Artificial Intelligence” by David Blumenthal, M.D., M.P.P.
October 22, 2025 at 1:15 PM
ai.nejm.org
The author of a letter proposes extending an AI-driven opportunistic screening model for coronary calcium to noncontrast head CTs to predict stroke risk through automated detection and quantification of intracranial arterial calcification. Learn more: nejm.ai/4ntZA3G

#AI #MedSky #MLSky
Page 1 of the letter "Extending Opportunistic AI Calcium Screening to Noncontrast Head CT for Stroke Risk Stratification" Bottom: Read full article at ai.nejm.org.
October 21, 2025 at 1:31 PM
ai.nejm.org
In a new Perspective, Daniel Restrepo, MD, argues that although GPT-4 demonstrates impressive diagnostic acumen, clinical diagnosis remains deeply embedded in the elements that current AI lacks. Read the full Perspective: nejm.ai/46W1Fzi

#AI #MedSky #MLSky
“Overreliance on [large language models] for reasoning carries a very real possibility of atrophying extant diagnostic reasoning skills as we increasingly outsource the aging practice of diagnosis.” Perspective “Raging against the Machine: The Human Side of the Story” by Daniel Restrepo, M.D.
October 20, 2025 at 12:30 PM
ai.nejm.org
In the latest episode of the NEJM AI Grand Rounds podcast, Dr. Jonathan Chen discusses his path from teenage programmer to Stanford physician-informatician and why machine learning has both thrilled and unnerved him. Listen to the full episode: nejm.ai/ep35

#AI #MedSky
AI GRAND ROUNDS EPISODE 35 Medicine, Machines, and Magic: Dr. Jonathan Chen on Medical AI
October 15, 2025 at 1:40 PM
ai.nejm.org
Editorial by Robert D. Truog, MD, MA, and R. Sean Morrison, MD: Machine Learning Cannot Replace Surrogate Decision-Makers in Resuscitation Decisions for Incapacitated Patients nejm.ai/3Vysymr

#AI #MedSky #MLSky
Page 1 of the editorial "Machine Learning Cannot Replace Surrogate Decision-Makers in Resuscitation Decisions for Incapacitated Patients" Bottom: Read full article at ai.nejm.org.
October 14, 2025 at 4:31 PM
ai.nejm.org
Original Article by G. Starke et al.: Machine Learning–Based Patient Preference Prediction: A Proof of Concept nejm.ai/3IAIYrD

#AI #MedSky #MLSky
Table 2. Characteristics of the 1811 Included Study Participants, Meaning All Participants for Whom the Target Variable Was Available. Table 3. Accuracy of Predicting Preferences Toward Attempting Cardiopulmonary Resuscitation in the Case of Cardiorespiratory Arrest among Couples in the Sample.
October 14, 2025 at 1:30 PM
ai.nejm.org
What does a chief health AI officer do? Dr. Karandeep Singh shares on NEJM AI Grand Rounds: Keep #AI smart, safe, and ethical in health systems. Learn more: nejm.ai/ep34

#MedSky #MLSky
October 13, 2025 at 12:31 PM
ai.nejm.org
Editorial by Issa J. Dahabreh, MD, ScD, Robert W. Yeh, MD, MSc, MBA, and Piersilvio De Bartolomeis, MSc: Trial Emulation, Simulation, and Augmentation Using Electronic Health Records and Generative AI nejm.ai/4mqiyXj

#AI #MedSky #MLSky
Page 1 of the editorial "Trial Emulation, Simulation, and Augmentation Using Electronic Health Records and Generative AI" Sitting at the bottom is "Read full article at ai.nejm.org."
October 10, 2025 at 3:32 PM