My research lies at the intersection of machine learning and healthcare. The overall goal is to use data and machine learning to make healthcare decisions smarter and more efficient.

Below are some highlights of my research, but for a complete accounting please see my google scholar profile page


Embeddings of clinical concepts


Paper: Beam, A.L., Kompa, B., Fried, I., Palmer, N.P., Shi, X., Cai, T. and Kohane, I.S., 2018. Clinical Concept Embeddings Learned from Massive Sources of Medical Data. arXiv preprint arXiv:1804.01486.
Link: Available at arXiv
Pretrained embeddings: Hosted at this link
Interactive explorer: Available here

Adversarial Attacks Against Medical Systems


Paper: Finlayson, S.G., Kohane, I.S. and Beam, A.L., 2018. Adversarial Attacks Against Medical Deep Learning Systems. arXiv preprint arXiv:1804.05296.
Link: Available at arXiv


The Cost of 17-Alpha Hydroxyprogesterone Caproat

Drug pricing in the United States is weird (some might say stupid), and often the cost of a drug is completely divorced from its value. The fallout due to the extreme price increases of epipen and Martin Shkreli’s daraprim have caused many to question the pricing practices of many drugs in the US. We examined one such instance of extreme pricing enabled by an unintended consequence of a piece of legislation known as the Orphan Drug Act, which led to a huge increase in price for 17P, a drug commonly used to prevent preterm birth. Our hope is that we can help shift the US healthcare debate from who has to pay for healthcare? to why is it so expensive in the first place? 17P

Paper: Fried*, I., Beam*, A.L., Kohane, I.S. and Palmer, N.P., 2017. Utilization, cost, and outcome of branded vs compounded 17-alpha hydroxyprogesterone caproate in prevention of preterm birth. JAMA internal medicine, 177(11), pp.1689-1690.
Link: Available at JAMA Internal Medicine
Background: More context can be found in this blog post
* Denotes equal contribution