Machine Learning and Artificial Intelligence: Why Should HIM Care?

By Susan H. Fenton, PhD, RHIA, CPHI, FAHIMA

It seems like every week brings a new newspaper article, journal article, or press release related to the expected impact of machine learning (ML) and artificial intelligence (AI) in health care.  On March 22, 2018, Daniel Faggella, writing for the website techemergence.com (https://www.techemergence.com/machine-learning-healthcare-applications/), enumerated the following current and future uses in healthcare:

CURRENT

  1. Diagnosing in Medical Imaging – computers read images better than we do
  2. Treatment Queries/Suggestions – truly advanced clinical decision support
  3. Drug Discovery
  4. Robotic Surgery

FUTURE

  1. Personalized Medicine – for drug efficacy or dosing
  2. Automatic Treatment – no human intervention required
  3. Improving Human Performance – beyond a simple activity tracker

You will notice that none of these is in the HIM field. However, Nuance now has their CAPD (computer-assisted physician documentation) tool that prompts physicians in real-time to improve their documentation. If you’ve worked seriously with data, you know that real-time correction is always preferred over retrospective. A little more searching could turn up other examples.

ML and AI are not without their challenges which includes, among other issues, the notion of trusting the “black box.” That is, many humans do not understand now ML and AI work.  How do you know you can trust it?

WHAT ABOUT HIM?

Health information managers need to be aware of the developments in ML and AI.  Further, for job security, they MUST project new developments over the next few years and be sure they can either add to the technology or identify the enhancements that human brings can to the problem.  

Join Dr. Susan Fenton at the MaHIMA Annual Conference on June 11, 2018 to learn more about ML and AI and its impact on HIM.