Uploaded August 2025 | Updated September 2026, 2 weeks ago
Models, Inference and Algorithms
May 28, 2025
Broad Institute of MIT and Harvard
Faisal Mahmood
Harvard Division of Medical Sciences
Advances in digital pathology and artificial intelligence have presented the potential to build models for objective diagnosis, prognosis and therapeutic-response and resistance prediction. In this talk we will discuss our work on: (1) Data-efficient methods for weakly-supervised whole slide classification with examples in cancer diagnosis and subtyping (Nature BME, 2021), identifying origins for cancers of unknown primary (Nature, 2021) and allograft rejection (Nature Medicine, 2022) (2) Discovering integrative histology-genomic prognostic markers via interpretable multimodal deep learning (Cancer Cell, 2022; IEEE TMI, 2020; ICCV, 2021; CVPR, 2024; ICML, 2024). (3) Building unimodal and multimodal foundation models for pathology, contrasting with language and genomics (Nature Medicine, 2024a, Nature Medicine 2024b, CVPR 2024). (4) Developing a universal multimodal generative co-pilot and chatbot for pathology (Nature, 2024). (5) 3D Computational Pathology (Cell, 2024) (6) Bias and fairness in computational pathology algorithms (Nature Medicine, 2024; Nature BME 2023) (7) Agentic AI workflows for diagnostic pathology and biomedical research.
Lab: mahmoodlab.org
Code: github.com/mahmoodlab
Key Articles
PathChat (Nature, 2024)
nature.com/articles/s41586-024-07618-3
TriPath (Cell, 2024)
cell.com/cell/fulltext/S0092-8674(24)00351-9
TOAD (Nature, 2021)
nature.com/articles/s41586-021-03512-4
CLAM (Nature BME, 2021)
nature.com/articles/s41551-020-00682-w
UNI (Nature Medicine, 2024)
nature.com/articles/s41591-024-02857-3
CONCH (Nature Medicine, 2024)
nature.com/articles/s41591-024-02856-4
For more information, visit: broadinstitute.org
Copyright Broad Institute, 2025. All rights reserved.
Models, Inference and Algorithms
May 28, 2025
Broad Institute of MIT and Harvard
Faisal Mahmood
Harvard Division of Medical Sciences
Advances in digital pathology and artificial intelligence have presented the potential to build models for objective diagnosis, prognosis and therapeutic-response and resistance prediction. In this talk we will discuss our work on: (1) Data-efficient methods for weakly-supervised whole slide classification with examples in cancer diagnosis and subtyping (Nature BME, 2021), identifying origins for cancers of unknown primary (Nature, 2021) and allograft rejection (Nature Medicine, 2022) (2) Discovering integrative histology-genomic prognostic markers via interpretable multimodal deep learning (Cancer Cell, 2022; IEEE TMI, 2020; ICCV, 2021; CVPR, 2024; ICML, 2024). (3) Building unimodal and multimodal foundation models for pathology, contrasting with language and genomics (Nature Medicine, 2024a, Nature Medicine 2024b, CVPR 2024). (4) Developing a universal multimodal generative co-pilot and chatbot for pathology (Nature, 2024). (5) 3D Computational Pathology (Cell, 2024) (6) Bias and fairness in computational pathology algorithms (Nature Medicine, 2024; Nature BME 2023) (7) Agentic AI workflows for diagnostic pathology and biomedical research.
Lab: mahmoodlab.org
Code: github.com/mahmoodlab
Key Articles
PathChat (Nature, 2024)
nature.com/articles/s41586-024-07618-3
TriPath (Cell, 2024)
cell.com/cell/fulltext/S0092-8674(24)00351-9
TOAD (Nature, 2021)
nature.com/articles/s41586-021-03512-4
CLAM (Nature BME, 2021)
nature.com/articles/s41551-020-00682-w
UNI (Nature Medicine, 2024)
nature.com/articles/s41591-024-02857-3
CONCH (Nature Medicine, 2024)
nature.com/articles/s41591-024-02856-4
For more information, visit: broadinstitute.org
Copyright Broad Institute, 2025. All rights reserved.
