illustration of DNA strand and abstract shapes on a computer screen on a green gradient background

Medical, health, and genomics

Digital pathology helps decode tumor microenvironments for precision immunotherapy. GigaPath is a novel vision transformer that can scale to gigapixel whole-slide images by adapting dilated attention for digital pathology. In joint work with Providence and UW, we’re sharing Prov-GigaPath, the first whole-slide pathology foundation model pretrained on large-scale real-world data, for advancing clinical research and discovery.

GigaPath: Whole-Slide Foundation Model for Digital Pathology

Research Forum Ep2 | Panel | Transforming the Natural Sciences with AI

Research Forum 2 | Panel Discussion: Transforming the Natural Sciences with AI

Scaling early detection of esophageal cancer with AI

female doctor / nurse reviewing a tablet

Microsoft Health Futures

Current selections

Sort by: Most recent

Clear selections

Search within these results

Content Types

Published Date

Publication

Improved estimation of the relationship between fetal growth and late stillbirth in the United States, 2014–15

Darren Tanner, Juan M. Lavista Ferres, Edwin A. Mitchell

Scientific Reports | March 2024

Project

Video

GigaPath: Foundation Model for Digital Pathology

March 11, 2024 | Naoto Usuyama

Presented by Naoto Usuyama at Microsoft Research Forum, Season 1, Episode 2 Naoto Usuyama proposed GigaPath, a novel approach for training large vision transformers for gigapixel pathology images, utilizing a diverse real-world cancer patient dataset, with the goal…

05:06

Video

Panel: Transforming the Natural Sciences with AI

March 11, 2024 | Bonnie Kruft, Alex Lu, Tristan Naumann, Kristen Severson, Tian Xie, Rianne van den Berg

Hosted by Bonnie Kruft, with Rianne van den Berg, Tian Xie, Tristan Naumann, Kristen Severson, and Alex Lu at Microsoft Research Forum, Season 1, Episode 2 Microsoft researchers shared their advancements in the fields of foundations models, drug…

40:43

Microsoft Research Blog

Scaling early detection of esophageal cancer with AI

March 11, 2024 | Anton Schwaighofer, Javier Alvarez-Valle

Microsoft Research and Cyted have collaborated to build novel AI models (opens in new tab) to scale the early detection of esophageal cancer. The AI-supported methods demonstrated the same diagnostic performance as the existing manual…

Publication

Enabling large-scale screening of Barrett’s esophagus using weakly supervised deep learning in histopathology

Kenza Bouzid, Harshita Sharma, Sarah Killcoyne, Daniel Coelho de Castro, Anton Schwaighofer, Maximilian Ilse, Valentina Salvatelli, Ozan Oktay, Sumanth Murthy, Lucas Bordeaux, Luiza Moore, Maria O'Donovan, Anja Thieme, Aditya Nori, Marcel Gehrung, Javier Alvarez-Valle

Nature Communications | March 2024, Vol 15

Publication

Towards a clinically accessible radiology foundation model: open-access and lightweight, with automated evaluation

Juan Manuel Zambrano Chaves, Shih-Cheng Huang, Yanbo Xu, Hanwen Xu, Naoto Usuyama (naotous), Sheng Zhang, Fei Wang, Yujia Xie, Mahmoud Khademi, Ziyi Yang, Hany Hassan Awadalla, Julia Gong, Houdong Hu, Jianwei Yang, Chunyuan Li, Jianfeng Gao, Yu Gu, Cliff Wong, Mu-Hsin Wei, Tristan Naumann, Muhao Chen, Matthew P Lungren, Serena Yeung-Levy, Curtis P. Langlotz, Sheng Wang, Hoifung Poon

March 2024

Microsoft Research Blog

ViSNet: A general molecular geometry modeling framework for predicting molecular properties and simulating molecular dynamics

February 29, 2024 | Tong Wang, Bin Shao, Tie-Yan Liu

Molecular geometry modeling is a powerful tool for understanding the intricate relationships between molecular structure and biological activity – a field known as structure-activity relationships (SAR). The main premise of SAR is that the biological…

The general model architecture of ViSNet. (a) Model sketch of ViSNet. ViSNet embeds the 3D structures of molecules and extracts the geometric information through a series of ViSNet blocks and outputs the molecule properties such as energy, forces, and HOMO-LUMO gap through an output block. (b) Flowchart of one ViSNet Block. One ViSNet block consists of two modules: i) Scalar2Vec, responsible for attaching scalar embeddings to vectors.; ii) Vec2Scalar. The inputs of Scalar2Vec are the node embedding, edge embedding, direction unit and the relative positions between two atoms.

Microsoft Research Blog

Research Focus: Week of February 19, 2024

February 21, 2024 | Anna Pavlenko, Joyce Cahoon, Yiwen Zhu, Brian Kroth, Jesús Camacho Rodríguez, Karla Saur, Sudipta Sinha, Stephen Schneider, Serena Hillman

In this issue: CaaSPER: vertical autoscaling algorithm dynamically maintains optimal CPU utilization; Improved scene landmark detection for camera localization runs faster, uses less storage; ESUS simplifies usability questionnaires for technical products and services.

Research Focus Week of February 19, 2024

Publication

Scalable nonparametric clustering with unified marker gene selection for single-cell RNA-seq data

Chibuikem Nwizu, Madeline Hughes, Michelle L. Ramseier, Andrew W. Navia, Alex K. Shalek, Nicolo Fusi, Srivatsan Raghavan, Peter S. Winter, Ava P. Amini, Lorin Crawford

bioRxiv | February 2024

Project

Publication

Self-Supervised Learning for Improved Optical Coherence Tomography Detection of Macular Telangiectasia Type 2

Shahrzad Gholami, Lea Scheppke, Meghana Kshirsagar, Yue Wu, Rahul Dodhia, Roberto Bonelli, Irene Leung, Ferenc B. Sallo, Alyson Muldrew, Catherine Jamison, Tunde Peto, Juan M. Lavista Ferres, Bill Weeks, Martin Friedlander, Aaron Y. Lee

JAMA Ophthalmology | February 2024

Project

Medical, health, and genomics

Highlights