AI (Artificial Intelligence) in EMR Software

Today I had an interesting conversation with MEDENT. It’s an EHR company that’s in only 8 states. I could actually write a whole post on just their approach to EMR software and the EMR market in general. They take a pretty unique approach to the market. They’ve exercised some restraint in their approach that I haven’t seen from many other EHR vendors. I’ll be interested to see how that plays out for them.

Their market approach aside, I was really intrigued by their approach to dealing with ICD-10. They actually described their approach to ICD-10 similar to how Google handled search. There’s all this information out there (or you could say all these new codes) and so they wanted to build a simple interface that would be able to easily and naturally filter the information (or codes in this case). A unique way of looking at the challenge of so many new ICD-10 codes.

However, that was just the base use case, but didn’t include what I consider applying AI (Artificial Intelligence) to really improve a user interface. The simple example they gave had to do with collecting data from their users about which things they typed and which codes they actually selected. This real time feedback is then added to the algorithm to improve how quickly you can get to the code you’re actually trying to find.

One interesting thing about incorporating this feedback from actual user experiences is that you could even create a customized personal experience in the EMR. In fact, that’s basically what Google has done with search with their search personalization (ie. when you’re logged in it knows your search history and details so it can personalize the search results for you). Although, when you start personalizing, you still have to make sure that the out of box experience is good. Plus, in healthcare you could do some great personalization around specialties as well that could be really beneficial.

I’d heard something similar from NextGen at the user group meeting applied to coding. The idea of tracking user behavior and incorporating those behaviors into the intelligence of the EMR is a fascinating subject to me. I just wonder how many other places in an EMR these same principles can apply.

I see these types of movements as part of the larger move to “Smart EMR Software.”

About the author

John Lynn

John Lynn

John Lynn is the Founder of, a network of leading Healthcare IT resources. The flagship blog, Healthcare IT Today, contains over 13,000 articles with over half of the articles written by John. These EMR and Healthcare IT related articles have been viewed over 20 million times.

John manages Healthcare IT Central, the leading career Health IT job board. He also organizes the first of its kind conference and community focused on healthcare marketing, Healthcare and IT Marketing Conference, and a healthcare IT conference,, focused on practical healthcare IT innovation. John is an advisor to multiple healthcare IT companies. John is highly involved in social media, and in addition to his blogs can be found on Twitter: @techguy.


  • Interesting post. I am a regular reader and am curious to know more about the restraint that you mentioned Medent is taking. I work for a number of practices that operate using their products, so I am curious to know more about how you feel about their posture.

  • these kinds of features are nice, but i think they make one crucial mistake: the keyboard. most people are terrible at typing. i don’t have any statistics, but i would venture a guess that typing proficiency among healthcare professionals is significantly below average.

    the answer to the keyboard problem, as best as we know today, can be click-based or voice based. of course, navigating 68,000 ICD codes via click is not particularly compelling. but there are good solutions to this problem. most physicians don’t use more than 30-40 codes on a regular basis, and it can be pretty easy to navigate those codes.

  • It’s true that the ability of an ICD code search AI to ‘remember’ your searching habits makes it all easier. The EMR we use (Claydata’s PuttyHealth) incorporates this kind of intelligent behaviour with Google-style instant searching – easy, even with my below average typing 😉

Click here to post a comment