Lightnews — Scholar-powered news

Curt Langlotz

@curtlanglotz.bsky.social

1.1K followers 370 following 52 posts

Director @StanfordAIMI, Sr Fellow @stanfordHAI, radiologist, machine learning geek, @RSNA past president, @StanfordBMI alum, author of The Radiology Report

Posts Media Videos Starter Packs

Pinned

Curt Langlotz @curtlanglotz.bsky.social · Dec 7

Two trivial critiques of this excellent story by the great @brian-casey.bsky.social: (1) due to RSNA succession rules, it was clear I would be president this year ever since I was elected to the board in 2016. (2) I don’t consider HIMSS a medical meeting; IMHO it’s an IT meeting. 🤓

Brian Casey @brian-casey.bsky.social · Dec 6

By reporting #RSNA24 attendance of 40k at mid-week (up 18% over last year), the meeting appears to have pulled ahead of #HIMSS24 (35k) for the title of largest US medical meeting. #HIMSS23 claimed the title in 2023 (35k vs. 34k for #RSNA23). buff.ly/4f7RrN4

RSNA Goes All-In on AI | RSNA Attendance Up 18% - The Imaging Wire

RSNA goes all-in on AI, mammo AI pays off, patient prep speeds MRI, opportunistic CT screening, imaging costs still scare patients, and more.

buff.ly

6 2 29

Reposted by Curt Langlotz

Tan-Lucien Mohammed, MD, FACR @tlhm-md.bsky.social · 16d

Radiologists are seeing significant reductions in radiograph reading times with the use of artificial intelligence-generated reports 📖

#radiology

@rasushrestha.bsky.social
@jimrawsonmd.bsky.social
@woojinkim.com

radiologybusiness.com/topics/artif...

Radiologists see significant reduction in reading times with AI-generated reports

Acceptability also increased considerably, “suggesting growing confidence in AI-generated reports over time,” experts write in JACR.

radiologybusiness.com

2 2

Curt Langlotz @curtlanglotz.bsky.social · 25d

👇

Woojin Kim, MD @woojinkim.com · 26d

Magical Thinking on AI

Friedman is not wrong to worry about what's going to happen vis a vis the U.S., China, and AI. However, we need less magical thinking and more realism.

aiguide.substack.com/p/magical-th...

Magical Thinking on AI

A Response to Thomas Friedman's Recent AI Columns in the New York Times

aiguide.substack.com

1 4

Reposted by Curt Langlotz

Stanford HAI @stanfordhai.bsky.social · Sep 8

In this Stanford news article, @stanfordhai.bsky.social Senior Fellow @suryaganguli.bsky.social argues that universities are essential for understanding how AI works. Here, he outlines three reasons why: news.stanford.edu/stories/2025...

Three reasons why universities are crucial for understanding AI

There is a “fierce urgency” to understand how artificial intelligence works, says Stanford physicist Surya Ganguli, who is leading a project to bring the inner workings of AI to light through transpar...

news.stanford.edu

3 11

Reposted by Curt Langlotz

Stanford HAI @stanfordhai.bsky.social · Sep 3

📸 Today from DC: @stanfordhai.bsky.social Faculty Affiliate Michelle Mello testified in Congress on AI in healthcare. To boost AI adoption, she discussed policy changes that could build trust in AI's performance. Read her testimony here: bit.ly/4ngI56k

1 5

Reposted by Curt Langlotz

Eric Topol @erictopol.bsky.social · Aug 27

Good to see @bsky.app has emerged as the social media platform of choice for science stuff
arstechnica.com/science/2025...

Bluesky now platform of choice for science community

It’s not just you. Survey says: “Twitter sucks now and all the cool kids are moving to Bluesky”…

arstechnica.com

17 310 1.2K

Reposted by Curt Langlotz

Dr Rasu Shrestha @rasushrestha.bsky.social · Sep 1

The phrase "artificial Intelligence" was born exactly 70 years ago.

Wow. 🤯

The phrase "artificial Intelligence" was born on this day 70 years ago.

The phrase "artificial intelligence" was coined by computer scientist John McCarthy in 1955. He used the term in a proposal for the 1956 Dartmouth Summer Research Project on Artificial Intelligence, a workshop that is widely considered the founding event of the field.

About the Dartmouth workshop
The proposal: In 1955, McCarthy, along with Marvin Minsky, Nathaniel Rochester, and Claude Shannon, proposed the idea that "every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it".
The event: The two-month workshop took place in the summer of 1956 at Dartmouth College.
The attendees: The workshop brought together key researchers who became leaders in early AI research and discussed how machines could use language, form abstractions, solve problems typically done by humans, and improve themselves.
The legacy: The Dartmouth workshop solidified AI as an academic field and set its initial goals.

1 1 3

Reposted by Curt Langlotz

Eric Topol @erictopol.bsky.social · Aug 30

Some generative A.I. tips for patients
www.wsj.com/opinion/a-ch...

15 21 65

Reposted by Curt Langlotz

Stanford HAI @stanfordhai.bsky.social · Aug 18

📣 Announcing the AI for Organizations Grand Challenge, a new competition for scholars to help organizations enter the era of AI. GoogleDeep Mind and @stanfordhai.bsky.social invite researchers from any university worldwide to submit your boldest ideas. Learn more: hai.stanford.edu/aiogc

1 4 8

Reposted by Curt Langlotz

Stanford HAI @stanfordhai.bsky.social · Aug 18

Congratulations to HAI Senior Fellow @suryaganguli.bsky.social on leading the new @simonsfoundation.org Collaboration on the Physics of Learning and Neural Computation! The collaboration will study the fundamental scientific principles underlying AI. www.simonsfoundation.org/2025/08/18/s...

2 11

Reposted by Curt Langlotz

Stanford HAI @stanfordhai.bsky.social · Aug 19

RadGPT helps patients understand their radiology reports. “We hope that our technology won’t just help to explain the results, but will also help to improve the communication between doctor and patient,” said @curtlanglotz.bsky.social, senior author of the study. hai.stanford.edu/news/new-lar...

New Large Language Model Helps Patients Understand Their Radiology Reports | Stanford HAI

‘RadGPT’ cuts through medical jargon to answer common patient questions.

hai.stanford.edu

1 8

Curt Langlotz @curtlanglotz.bsky.social · Aug 13

Medicine needs Neuro-Bayesian AI (paywall): www.wsj.com/articles/mee...

www.wsj.com

1 3

Reposted by Curt Langlotz

Brian Casey @brian-casey.bsky.social · Jul 11

Good timing too as the FDA yesterday released its updated list of AI-enabled medical devices that have received marketing authorization.

1 1

Curt Langlotz @curtlanglotz.bsky.social · Jul 11

Such important work! www.nature.com/articles/s41...

How AI is used in FDA-authorized medical devices: a taxonomy across 1,016 authorizations - npj Digital Medicine

npj Digital Medicine - How AI is used in FDA-authorized medical devices: a taxonomy across 1,016 authorizations

www.nature.com

3 7

Reposted by Curt Langlotz

Daniel Coelho de Castro @dccastr0.bsky.social · Jun 26

📣 Our new PadChest-GR benchmark in collaboration with @ua.es and MedBravo is published in @ai.nejm.org: ai.nejm.org/doi/full/10....!
📰 Also check out our blog post for why we're so excited about it: www.microsoft.com/en-us/resear...
💾 The dataset can be downloaded here: bimcv.cipf.es/bimcv-projec...

2 6

Curt Langlotz @curtlanglotz.bsky.social · Jul 10

Oops

Aneesh Sathe @aneeshsathe.com · Jul 9

This is funny

retractionwatch.com/2025/06/30/s...

Springer Nature book on machine learning is full of made-up citations

Would you pay $169 for an introductory ebook on machine learning with citations that appear to be made up? If not, you might want to pass on purchasing Mastering Machine Learning: From Basics to Ad…

retractionwatch.com

Reposted by Curt Langlotz

Stanford HAI @stanfordhai.bsky.social · Jul 3

A new large language model called “RadGPT” can help patients understand their radiology reports. Learn how a team of Stanford researchers developed the model: hai.stanford.edu/news/new-lar...

New Large Language Model Helps Patients Understand Their Radiology Reports | Stanford HAI

‘RadGPT’ cuts through medical jargon to answer common patient questions.

hai.stanford.edu

2 4

Reposted by Curt Langlotz

Mustafa Suleyman @mustafasuleymanai.bsky.social · Jun 30

We're taking a big step towards medical superintelligence. AI models have aced multiple choice medical exams – but real patients don’t come with ABC answer options. Now MAI-DxO can solve some of the world’s toughest open-ended cases with higher accuracy and lower costs.

2 6 17

Reposted by Curt Langlotz

Mustafa Suleyman @mustafasuleymanai.bsky.social · Jun 30

This research is just the first step on a long, exciting journey. We’re excited to keep testing and learning with our healthcare partners in pursuit of better, more accessible care for people everywhere. More on the blog today: microsoft.ai/new/the-path...

The Path to Medical Superintelligence | Microsoft AI

microsoft.ai

2 6

Reposted by Curt Langlotz

NEJM.org @nejm.org · Jun 25

The National Academy of Medicine Shared Commitments provide common-ground expectations for the integrative benefits of learning health systems, which align science, incentives, and culture to facilitate continuous improvement. Full Sounding Board article: nej.md/44js5Zv

@nam.edu #MedSky

A diagram illustrating the learning health system at work

4 3

Reposted by Curt Langlotz

Eric Topol @erictopol.bsky.social · Jun 16

The promise and problems with multi-cancer early detection tests
by Sid Mukherjee @newyorker.com
www.newyorker.com/magazine/202...

The Catch in Catching Cancer Early

New blood tests promise to detect malignancies before they’ve spread. But proving that these tests actually improve outcomes remains a stubborn challenge.

www.newyorker.com

3 36 110

Reposted by Curt Langlotz

Woojin Kim, MD @woojinkim.com · Jun 14

FDA issues final order reclassifying radiological CAD software

www.auntminnie.com/imaging-info...

FDA issues final order reclassifying radiological CAD software

Computer-assisted detection (CAD) and diagnosis algorithms used in radiology move from the class III to the class II category.

www.auntminnie.com

1 1

Curt Langlotz @curtlanglotz.bsky.social · Jun 11

Obviously this study doesn’t prove that point. I should have said “maybe”…

Curt Langlotz @curtlanglotz.bsky.social · Jun 11

Obviously this study doesn’t prove that point. I should have said “maybe”…

Curt Langlotz @curtlanglotz.bsky.social · Jun 10

Congratulations to @erichorvitz.bsky.social and team—such nice work. AI + human is still better than either one alone if the collaboration is well designed.

Eric Horvitz @erichorvitz.bsky.social · Jun 10

Study on clinician-AI collaboration: "As LLMs now demonstrate expert-level diagnostic performance, the focus shifts from whether AI can offer valuable suggestions to how it can be effectively integrated into the diagnostic workflow..." Many opportunities ahead.

www.medrxiv.org/content/10.1...

1 1

Curt Langlotz @curtlanglotz.bsky.social · Jun 10

Congratulations to the authors—such nice work. AI + human is still better that either one alone if the collaboration is well designed.

Ethan Mollick @emollick.bsky.social · Jun 9

New RCT shows a familiar result on LLMs & medicine:

Doctors given clinical vignettes produce significantly more accurate diagnoses when they also consult with a custom GPT built with the (obsolete) GPT-4 than doctors with Google/Pubmed but not AI.

Yet AI alone is as accurate as doctors + AI.

1 3