Very cool stuff from Faisal Mahmood group at Harvard Medical School Brigham and Women's Hospital pioneering work with #AI and #pathology, here in nature

LAB-Bench is, to my knowledge, the first set of evals designed to measure whether models/agents can do scientific research, not just whether they know assorted trivia about science. Procedural evaluations like this for complex tasks will be very important going forward.

Interesting 🧵from Max Lu Richard J. Chen showcasing how the community is using our pathology foundation models UNI (nature.com/articles/s4159…) and CONCH (nature.com/articles/s4159…).

Congratulations 🎉 to Dr. Sharifa Sahai for completing her PhD HMS_SysBio, and special thanks to everyone on the committee Debora Marks Nasir Rajpoot Georg K Gerber.

If you're at #ICML2024 ICML Conference come check out our work on multimodal prototypes and say hi to our superstar Andrew H. Song Paper: arxiv.org/pdf/2407.00224 Poster #202: Multimodal Prototyping for cancer survival prediction When/Where: 7/24 (Wed) 11:30AM ~ 1PM CEST Hall C 4-9

Research Highlight: Vision–language AI assistance in human pathology nature.com/articles/s4158…

Happy to share a new computational pathology preprint from katherlab & many other groups: "HIBRID - Histology and ct-DNA based Risk-stratification with Deep Learning". Deep Learning 💻+ ctDNA 🧬= very powerful prognostic model in colorectal cancer ☑️ medrxiv.org/cgi/content/sh…

⚡️🔬📣 Here are our two latest preprints on how AI for Pathology can advance pre-clinical drug safety and toxicity assessment. Work led by our superstar postdoc Guillaume Jaume: Deep Learning-based Modeling for Preclinical Drug Safety Assessment 📄 Preprint:

The pace of progress in GenAI for pathology has been stunning but, unfortunately, there are major obstacles for it to move into routine medical care. Faisal Mahmood takes us through this in the new Ground Truths podcast, with transcript, #AI open-access, link in profile

📣 I'm thrilled to share that I’ll be joining MIT as an assistant professor this fall, joint between MIT Media Lab & MIT EECS. My group will advance the foundations of multisensory AI to enhance the human experience. I look forward to tackling exciting challenges in multimodal AI

3D tissues can reveal more about disease than traditional 2D slices, but they are enormously complex. Find out how a new AI tool can analyze these data-rich specimens to predict outcomes: nibib.nih.gov/news-events/ne… Andrew H. Song Jonathan Liu Faisal Mahmood Mechanical Engineering at UW

Excited to work on this with Alpenglow Biosciences Faisal Mahmood Eben L Rosenthal, MD Michael Topf MD, MSCI and our talented engineers and clinicians University of Washington !

⚡️🔬📣 We are excited to announce our new #ECCV 2024 European Conference on Computer Vision #ECCV2024 paper "Multistain Pretraining for Slide Representation Learning in Pathology" Led by Guillaume Jaume & Anurag Vaidya this work is the latest iteration of our efforts on whole slide representation learning for

The 5 different patterns of brain aging from ~50,000 individuals, with their relationship to lifestyle, genetics and outcomes, derived from MRI and #AI. nature.com/articles/s4159… Nature Medicine Christos Davatzikos and colleagues

This article (arxiv.org/pdf/2408.15823) from Jakob Nikolas Kather is the most comprehensive effort to compare comp path foundation models to date. Showing that multimodal FMs like CONCH outperform unimodal FMs on unimodal tasks, and that the diversity of data >> quantity of data.

Thesis defended and officially accepted! 🎓✨ Grateful for the support of my family, friends, community, and lab throughout this journey at Harvard. Onward to the next chapter! #PhDone

A very thoughtful essay on #AI and the practice of medicine, a moral endeavor that can be tech supplemented but needs a "hard boundary restricting its deployment for tasks requiring moral agency." jamanetwork.com/journals/jamac… by Sarah C. Hull, MD, MBE and Joseph Fins JAMA Cardiology

With Jana Lipkova we wrote about "The age of foundation models in pathology" for NatureRevClinOncol, discussing the Virchow model from Thomas J. Fuchs's team & others 🔬 journal link: nature.com/articles/s4157… full text: rdcu.be/dS4JO