About
Adam Casson
I'm a research engineer at Paige.AI where I work on self-supervised learning for large-scale vision models as well as multimodal language models for gigapixel histopathology imagery. I've co-authored papers that appeared in Nature Medicine (opens in a new tab), Cancer Research (opens in a new tab), MIDL (opens in a new tab), and other venues. I was also a core contributor to the only FDA approved AI system (opens in a new tab) for cancer diagnosis in pathology.
Experience
Paige.AI, Senior Research Engineer 2019-Present
- Working on weakly supervised and self supervised neural networks applied to histopathology images (100,000px by 100,000px in size).
- Developed network architecture that improved detection and localization of cancer.
- Core contributor to the research and development of Paige Prostate which received the first ever FDA approval of an AI system in pathology (read the paper here).
- Worked on developing ML training framework to increase reproducibility, experiment tracking, code modularity, and lowering the barrier of entry to running experiments.
- Involved in hiring most of the AI team by helping shape the interview process and doing 100+ technical interviews.
- Helped lead research and development of a weakly supervised breast cancer detection system (paper coming soon).
Comcast-NBCUniversal, Machine Learning Engineer 2017-2019
- Lead research of temporal language modeling for understanding semantic drift.
- Worked on facial recognition, object detection, and scene detection for long-form videos.
- Organized and taught a weekly machine learning course for coworkers.
Rochester Institute of Technology, Research Assistant 2016-2017
- Researched visual question answering (VQA) on video data.
- Developed a multi-modal network to jointly reason over videos and natural language to answer questions about a given video.
- Created a new video VQA dataset by using existing video captioning dataset by converted captions to question-answer pairs.
- Developed a hand-crafted system using dependency parsing to generate diverse sets of question-answer pairs from captions.
Education
Rochester Institute of Technology, B.S. Imaging Science 2013-2017
Selected Publications
Virchow2: Scaling Self-Supervised Mixed Magnification Models in Pathology (opens in a new tab)
arXiv preprint (arXiv:2408.00738) 2024
Zimmermann, E., Vorontsov, E., Viret, J., Casson, A., Zelechowski, M., Shaikovski, G., Tenenholtz, N., Hall, J., Fuchs, T., Fusi, N., Liu, S., Severson, K.
Nature Medicine 2024
Vorontsov, E.†, Bozkurt, A.†, Casson, A.†, Shaikovski, G.†, Zelechowski, M.†, Severson, K.†, Zimmermann, E., Hall, J., Tenenholtz, N., Fusi, N., Yang, E., Mathieu, P., van Eck, A., Lee, D., Viret, J., Robert, E., Wang, Y.K., Kunz, J.D., Lee, M.C.H., Bernhard, J.H., Godrich, R.A., Oakley, G., Millar, E., Hanna, M., Wen, H., Retamero, J.A., Moye, W.A., Yousfi, R., Kanan, C., Klimstra, D.S., Rothrock, B., Liu, S., Fuchs, T.J.
† Equal contribution
PRISM: A Multi-Modal Generative Foundation Model for Slide-Level Histopathology (opens in a new tab)
arXiv preprint (arXiv:2405.10254) 2024
Shaikovski, G.†, Casson, A.†, Severson, K., Zimmermann, E., Wang, Y., Kunz, J.D., Retamero, J.A., Oakley, G., Klimstra, D., Kanan, C., Hanna, M., Zelechowski, M., Viret, J., Tenenholtz, N., Hall, J., Fusi, N., Yousfi, R., Hamilton, P., Moye, W.A., Voronstov, E., Liu, S., Fuchs, T.J.
† Equal contribution
Adadpting Self-Supervised Learning for Computational Pathology (opens in a new tab)
CVPR Workshop on Data Curation and Augmentation in Medical Imaging (DCAMI) 2024
Zimmermann, E., Tenenholtz, N., Hall, J.B., Shaikovski, G., Zelechowski, M., Casson, A., Milletari, F., Viret, J., Voronstov, E., Liu, S., Severson, K.A
Medical Imaging with Deep Learning (MIDL) 2023
(Oral presentation, MedIA special issue selectee)
Casson, A.†, Liu, S.†, Godrich, R.A., Aghdam, H., Lee, D., Rothrock, B., Kanan, C., Retamero, J., Hanna, M., Millar, E., Klimstra, D., Fuchs, T.
† Equal contribution
© Adam Casson.