Machine learning research as med student/resident

This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.

djp724

New Member
Joined
Jul 29, 2020
Messages
2
Reaction score
0
Hi, I’m currently one year out from college and am working as an applied ML researcher in industry. I’m starting the med school application process now but am having doubts about whether I’ll be able to use my technical background if I continue down the MD route. Although I certainly want to be a clinician, I also have a strong desire to continue doing ML research in academia or industry (preferably industry, most of my jobs/internships so far have been at startups and I’ve really appreciated the fast pace of those environments). Am I deceiving myself by thinking I’ll have opportunities to do both? Is this a jack of all trades, master of none situation? I’m not a PhD, but I do have an undergrad and masters from a top CS school, as well as a first author publication in a biostat journal.

Members don't see this ad.
 
If that’s your love, I would pursue that especially if you already have technical skill in that area. Medicine is a demanding field mainly commitment wise which makes it hard to pursue other interests like that on the side. There are several exceptions and maybe many of us here have them, but the general rule is other interests tend to take a backseat. It’s a 7-10 year slog of doing what your superiors tell you, after which you can then transition to something like machine learning if that interest is still intact. Let’s here what others say.
 
Last edited:
The MD route will not give you any further training in ML, so you won’t become an expert in the field the same way someone who pursues a PhD in it might. But if you already have skills in the area and/or can find people to mentor you, you may be able to find research opportunities that relate to it.

Not at all ML, but I’ve done some pretty non traditional research in med school so it’s possible if you seek it out. It just may take more work to find mentors and/or learn skills than in more traditional clinical research.
 
So ML research in medicine is huge right now (I’m using that term broadly). I’ve published some work using DCNN and countless other groups are doing similar things.

The biggest struggle is bridging the gap between the comp guys and the clinicians. We have questions but it’s often hard to get everyone to understand the nuance involved. An MD with a solid comp background would be very valuable. You could definitely find work and funding in academia, and there are infinite opportunities to develop IP and spin off startups from that position. I’m still learning that aspect of it, though do have some IP currently going through the institutional process. Many colleagues have little startups they’ve developed through their academic positions. Of course you do even up having to share your financial successes with your institution according to their IP policies, but they also provide a lot of funding and assume some of the early risk.
 
  • Like
Reactions: 1 users
Top