“Challenges and Opportunities for Machine Learning in Cancer Immunotherapy”

Presented by Jennifer Chayes, Managing Director, Microsoft New England Research and New York City

Talk Description: Cancer immunotherapy is one of the most exciting developments in healthcare. By enlisting our own immune systems to go after cancer, immunotherapy is much more focused than other cancer therapies – when effective, it kills the cancer cells without the damage of chemo or radiation. Cancer immunotherapy is now a first line treatment for stage four melanoma and lung cancer, leading to long-term survival in many patients. On the other hand, existing cancer immunotherapies only work for a relatively small subset of cancer patients. How do we properly identify the patients for whom existing immunotherapies are likely to be effective, and not have serious side effects, and how do we develop new cancer immunotherapies for other patients? The relevant data is multimodal: genomic, immunological, metabolic, clinical and more. Trials involve relatively few patients with extremely high-dimensional data per person. This poses unique challenges and opportunities for machine learning and statistics.

This video is part of IACS’s 2018 Symposium on the Future of Computation in Science and Engineering. This year our annual symposium focused on how medicine and health care are being reshaped by computational science, big data, and information technology.

To learn more, visit our website at