1/4 Interested in multimodal fusion in Computational Pathology? Checkout our preprint combining transcriptomics and histology for survival prediction! Paper: arxiv.org/abs/2304.06819 Code: github.com/ajv012/SurvPath Anurag Vaidya Richard J. Chen @DFKW_MD Paul Liang Faisal Mahmood

Join us at ICML'23 to discuss deployment challenges for Generative AI with this amazing set of invited speakers! CFP: tinyurl.com/deployinggener… Deadline: May 19th, 2023 Anywhere on Earth (AoE) Notifications: June 23rd, 2023 Workshop date: July 28, 2023

⚡️🔬📣Excited to share our two new Nature Medicine articles, we develop computational pathology foundation models, 1. UNI, a self-supervised computational pathology model trained on 100 million pathology images from 100k+ slides. 2. CONCH, a vision-language model for

.Raj Movva, Pang Wei Koh, and I write for Nature Medicine on using unlabeled data to improve generalization + fairness of medical AI models: nature.com/articles/s4159… We highlight two nice recent papers illustrating this - nature.com/articles/s4159…, nature.com/articles/s4159….

⚡️🔬📣Excited to share our new Nature Medicine article, examining disparities in pathology AI models, assessing how modeling choices impact disparities, and evaluating the potential of self-supervised foundation models in mitigating these disparities. nature.com/articles/s4159… See

Excited to have been part of this awesome effort! We investigate an important problem of disparity in AI models for computational pathology, by thoroughly sifting through several modeling strategies. Enjoy reading this timely piece!

Excited to share our latest publication in Nature Medicine! 🎉 Proud to have been part of this incredible team effort. We study the disparity and fairness in AI models for computational pathology, exploring a variety of modeling strategies. Check it out below! 👇

What an incredible year for Faisal Mahmood lab! Leading the field in computational pathology and now doing excellent work on disparities.

New studies suggest that using unlabelled data in medical #AI can improve accuracy and generalization to new settings and minority patient groups, thereby increasing fairness. News and Views from Emma Pierson & colleagues Emma Pierson Cornell Tech nature.com/articles/s4159…

Are you interested in working at the interface of machine learning and pathology? Consider joining @DFKW_MD's lab at Emory.

If you learn a Puli’s a dog breed, you’d correctly infer they’ve got four legs and fur. But when knowledge editors inject facts into LLMs, do they generalize too? Using our new TAXI dataset, we find editors indeed generalize, but way worse than people arxiv.org/abs/2404.15004

Based on numerous requests we are providing the open access ShareIT link for our Nature Medicine article on identifying and mitigating disparities in pathology AI models. Open read link: rdcu.be/dFdMS Journal link: nature.com/articles/s4159…

Excited to share our review on hemostatic agents, where we summarize trends in 54 approved hemostats and 75 active clinical trials. Check out this timely article - with Pfizer’s latest gene therapy for Hemophilia B approved just last week! Samir Mitragotri Zongmin Zhao @BioTM_Buzz

New papers out 📣⚡️! In SurvPath, we integrate cellular pathways into multimodal survival prediction. In TANGLE, we use transcriptomics to guide histology slide representations. Special shout out to Guillaume Jaume and rest of the team. Meet us at #CVPR2026 2024!

Using artificial intelligence, researchers from Mass General Brigham have created and trained a new model using 3D datasets to predict the recurrence of prostate cancer. The model outperformed other models that rely on 2D datasets. buff.ly/3VnW7YH Faisal Mahmood Andrew H. Song

🚨⚡️Super excited to be part of HEST-1K, bringing you over 1.5M histology-omics pairs. Have fun with this public dataset for representation learning, biomarker discovery, and benchmarking models! Massive efforts by Guillaume Jaume Paul Doucet Faisal Mahmood

🚨⚡️Excited to bring you MMP - a highly interpretable and computationally light method to do multimodal prognostication in computational pathology. Amazing effort by Andrew H. Song Richard J. Chen Guillaume Jaume and Faisal Mahmood

🚨🔬 Super thrilled to announce a fun project with Guillaume Jaume! Using 7,000+ slides and 20+ tasks, we show that aligning H&E and immunohistochemistry slides results in robust gigapixel slide encoders.

⚡️📣After the success of our previous pathology foundation models UNI (rdcu.be/dBMgh) and CONCH (rdcu.be/dBMf6), we are now announcing TITAN (arxiv.org/abs/2411.19666), a new state-of-the-art whole slide level foundation model trained on >330k pathology slides

✨📣Introducing THREADS: a multimodal foundation model for pathology trained on paired histology and genomic data 🔬+🧬 We show that: (a) THREADS achieves SOTA performance on >50 tasks in oncologic pathology with much less pre-training data than other models, highlighting the