Sumit Chopra

[intermediate] Deep Learning for Healthcare

Summary

Machine learning (ML), especially Deep Learning (DL), is expected to have a significant impact on healthcare, with a widely accepted belief that it will eventually transform the field. This course will provide a window into this exciting interdisciplinary field, both from a research and practical perspective. We will first set the stage by giving an overview of clinical care, describing the various datasets in healthcare, and talking about the unique challenges in this field. We will then discuss numerous ways in which deep learning has made an impact by solving problems using medical imaging, electronic health records (EHRs), and time series datasets, with models like convolutional neural networks, recurrent neural networks, and graph neural networks; and techniques like fully supervised learning, self-supervised learning, and reinforcement learning. We will then talk about future work in this interdisciplinary area and conclude by discussing the challenges associated with translating research into clinical practice.

Syllabus

• Overview of clinical care
• Datasets in healthcare
• Unique challenges associated with using machine learning/deep learning models in healthcare
• Medical image analysis and deep learning
• Deep learning for electronic health records (EHRs)
• Deep learning models for time series data in healthcare
• Deep reinforcement learning in healthcare
• Considerations in building an AI-based clinical device

References

Rajpurkar, P., Chen, E., Banerjee, O. et al. AI in Health and Medicine. Nat Med 28, 31–38 (2022).

Ghassemi M., Naumann T., Schulam P., Beam A.L., Chen I.Y., Ranganath R. A Review of Challenges and Opportunities in Machine Learning for Health. AMIA Jt Summits Transl Sci Proc. 2020 May 30;2020:191-200.

Miotto, R., Li, L., Kidd, B. et al. Deep Patient: An Unsupervised Representation to Predict the Future of Patients from the Electronic Health Records. Sci Rep 6, 26094 (2016).

Zhu, Weicheng, and Narges Razavian. Variationally regularized graph-based representation learning for electronic health records. Proceedings of the Conference on Health, Inference, and Learning. 2021.

Yiqiu Shen, Nan Wu, Jason Phang, Jungkyu Park, Kangning Liu, Sudarshini Tyagi, Laura Heacock, S. Gene Kim, Linda Moy, Kyunghyun Cho, Krzysztof J. Geras. An interpretable classifier for high-resolution breast cancer screening images utilizing weakly supervised localization. Medical Image Analysis, Volume 68, 2021, 101908, ISSN 1361-8415.

Kim M., Yun J., Cho Y., Shin K., Jang R., Bae H.J., Kim N. Deep Learning in Medical Imaging. Neurospine. 2019 Dec;16(4):657-668.

Lee, J., Yoon, W., Kim, S., Kim, D., Kim, S., So, C.H., Kang, J. (2020). BioBERT: a pre-trained biomedical language representation model for biomedical text mining. Bioinformatics, 36(4), 1234-1240.

Prasad, N., Cheng, L., Chivers, C., Draugelis, M., Engelhardt, B.E. (2017). A Reinforcement Learning Approach to Weaning of Mechanical Ventilation in Intensive Care Units. ArXiv, abs/1704.06300.

Pre-requisites

Familiarity with linear algebra, probability, calculus, and machine learning.

Short bio

Dr. Sumit Chopra is an Associate Professor at the Courant Institute of Mathematical Sciences, NYU. In addition, he holds a joint position of an Associate Professor in the Department of Radiology, NYU Grossman School of Medicine, where he is also the Director of Machine Learning Research. He received his MS and PhD degrees from Courant Institute of Mathematical Sciences, NYU. Prior to being a faculty at NYU, he was the co-founder of a well-funded startup called Imagen Technologies, developing AI-based medical devices. Before Imagen, he was a senior research scientist at Facebook AI Research (FAIR) (now Meta AI Research) and at AT&T Labs-Research. His research interests lie in statistical machine learning, focusing on representation learning in self-supervised settings, with applications in computer vision, natural language understanding, and healthcare. He has published more than 50 papers in the field and has led the development of multiple AI-based medical devices approved by the United States Food and Drug Administration (FDA).