Natural Language Processing in Healthcare - A Clinician's Perspective

Henry Feldman, M.D.



Healthcare data is heavily stored in narrative form of unstructured text. Unlike non-medical text, the language used in healthcare is often extremely complex, filled with complex negation, hypotheticals, and jargon, but at the same time using this data requires high-precision understanding. While the amount of text in a patient’s chart is small by computer standards, it is rapidly exceeding human capacity for comprehension; this makes the use of Natural Language Processing (NLP) a natural fit for advanced computation on healthcare notes.

Learning Objectives:
After attending this lecture, students and healthcare professionals will be able to:
- Define what Natural Language Processing is
- Understand various approaches to NLP and what the advantages of each are
- See some real-world use cases for NLP
- Learn about some of the standard ontologies in use to encode the output of the NLP
tools (UMLS)
- Understand some of the common pitfalls in comprehending medical text

Target audience:
Medical, Nursing, Pharmacy, Public health, Dental and Veterinary students, and clinicians in practice. NOTE: A higher-level discussion can be arranged/added with researchers in mind.

Aligns to:
For physicians and non-physicians, this lecture aligns to LCME and AGCME competencies of Medical Knowledge and Clinical Reasoning, Interpersonal and Communication Skills (AI/NLP).

For nursing, this lecture aligns to AACN Essential Core Competencies of Information and Technology, Academic-Practice Partnerships, Systems-based Practice, and Career-long Learning.

For pharmacists, this lecture supports learning objectives for ACPE Doctor of Pharmacy curriculum requirements for Clinical Sciences – Health Informatics and competencies for ASHP PGY1 Residencies and PGY2 Residencies in Pharmacy Informatics.

For veterinary medicine, this lecture aligns with the competencies of Problem Solving, Critical Thinking, and Life-long Learning and Ethics.

Conflict of Interest: Henry Joel Feldman, MD is employed by IBM Watson Health which develops and markets products and services across the healthcare sector. IBM is a manufacturer of regulated medical devices.


Photo of Henry Feldman, M.D.

Henry Feldman is an Assistant Professor of Medicine at Harvard Medical School,as well as an Assistant Professor of Clinical Science at the Cummings School of Veterinary Medicine at Tufts University. He is a board-certified physician in Internal Medicine/Focused Practice Hospital Medicine as well as Clinical Informatics.

Henry worked in the computer industry for 10 years, prior to going to medical school at the NYU School of Medicine. He completed an internship, residency, and chief residency in Internal Medicine at NYU/Bellevue Hospital in New York. He then completed a 2-year Medical Informatics fellowship at NYU School of Medicine.

Henry joined the faculty of Beth Israel Deaconess Medical Center (BIDMC) at Harvard Medical School in Boston as a Hospitalist in 2006. He served as the Chief Information Architect of the Division of Clinical Informatics at BIDMC, from2007-2018, where he managed the software design and development teams.

In May 2018 Henry joined IBM Watson Health as the Deputy Chief Medical Officer for Development, and he continues to practice medicine on the covered and uncovered hospital medicine service as a nocturnist at BIDMC. Most of his hospital-based research continues in clinical 3D printing, for surgical devices.

As an offshoot of this research, he co-formed a course at Harvard Medical School in the department of surgery for CAD/CAM for medical devices particularly for surge.

Lecture booking request

Thank you for your interest in hosting an IBM speaker. Please fill out the following form with as much detail as possible. An IBM representative will reach out to discuss your booking request. All guest lectures are subject to availability and agreements under this collaboration are not legally binding.