After a long career at GE, Shan Jegatheeswaran joined Johnson & Johnson last year as senior vice president of digital for the medical technology business. In this new role, Jegatheeswaran is building tools to improve patient care and help surgeons operate with greater precision and efficiency.
In an interview with MedTech Dive, he talked about transitioning from a career in energy to a job in healthcare, his first internship, and the tasks people don’t think much about building successful AI tools. .
This interview has been edited for length and clarity.
MEDTECH DIVE: When did you join J&J?
Jegatishwaran: I started in July. Prior to that, he was the Chief Digital Officer for Oilfield Services, a business unit of Baker Hughes. [which] It’s effectively a spin-off of GE Oil and Gas. I joined GE, worked in multiple businesses at GE, spent the last nine years in oil and gas, and prior to that in capital, healthcare, aviation, and corporate. It was a good experience.
What inspired you to return to medical work?
I love the healthcare opportunity set. You hear a lot about the industry when you’re outside of it, but it’s very personal to most people.
A little-known fact: J&J was my first internship. So my goal has always been to work in pharma/healthcare. Actually, I have an employee ID card from 20 years ago.
What kind of projects does J&J medtech work on?
My scope is basically software that works very closely with medtech devices, or software that enables experiences on medtech devices. One degree away from the OR, so to speak. Before surgery, during surgery, and after surgery. The scope of my team has to do with all that stuff. As you can imagine, our goal is to enable a frictionless and seamless experience across the board for patients, medical teams, and support teams.
Which technology are you most excited about when it comes to software in the medical technology industry?
What excites me the most is the people element of digital stories.It’s often overlooked…but what’s actually very interesting, and what drives success, is how you drive that intersection [the] human and digital experience. There is a little bit of art, but there is also a lot of science.
And AI is becoming more and more prevalent, right? My mother talks about it at dinner now. she is everywhere. So how do we imbue our everyday environment with AI so that it’s just part of what we do, not AI?
What needs to happen behind the scenes for AI to work well?
When I say AI, it looks like there are three boxes on the screen: data on the left, AI in the middle, and values on the right. And it’s not.
There are procedures around data aggregation, data tagging, data cataloging, data auditing, data policies, and even data management before we get into analytics. It’s fair to say a lot of sweat needs to be done. Many of them are manual, at least initially.
We’ve settled on the Netflix of the world and the Google of the world because they already have a tagged data set.
Shooting something like a video post-procedure for a surgical operation can sort of make a four-hour cut where the doctor wants a highlight reel. This is one of the factors that we have spent some time considering from an AI perspective.
It’s a blockage and a struggle around cataloging [and] Data aggregation. That’s why we work closely with partners like Microsoft to build these solutions in-house.
How do you prevent bias when developing algorithms?
I think first and foremost is the discipline of the control infrastructure regarding how the design is executed and the testing of the algorithms themselves.
We have guidelines on how to approach it across the pharmaceutical sector.we are leveraging some [those guidelines] We will also make it available to move to medical technology use cases.
From our point of view, it should be noted that we do not obtain homogenous datasets from the device, nor do we consider profiles of patients not serviced by the device. Fortunately, J&J has a very broad reach from a market perspective — we have a fairly diverse set of data — but we spend a lot of time with academia and government to We have confirmed that some of them have been suppressed.
Second, you’ve already tested analytics before you put them on the market that provide some guidance or insight.
Speed does not necessarily have to be the leading indicator in all cases. Quality should be. Therefore, we are fortunate to be able to build software that is competent and backed by the best talent from a regulatory, privacy, and governmental perspective.
