⁂ George Ho

“So What Is Your Job, Exactly?”

My team at Flatiron Health (but especially the inimitable Blythe Adamson) recently wrote an excellent overview of how we use machine learning to build research-ready oncology datasets from patient medical records — in a paper called Approach to Machine Learning for Extraction of Real-World Data Variables from Electronic Health Records.

This is a refreshingly simple and concise articulation of what I do at my day job! You can read the paper on Frontiers in Pharmacology, medRxiv, or read a brief overview on the Flatiron website.

To spare you a few clicks, here’s what I think is the most enlightening figure in the paper:

Snippets of text from electronic health records are inputs to deep learning models that produce a variable value for each patient as an output. Source: Figure 5 in the paper.

#Cancer #Machine-Learning #Natural-Language-Processing