AI PhD to Med School - Smart Move?

[PubPhD]

.

---

Members don't see this ad.

 

[Goro]

I don't see why you can't help people by continuing on the path you're won, and I also don't see how being an MD will add any benefit to your goals

 

[amuseddoughnut]

The only reason to get the MD is if you want to personally treat patients. Without an MD, no one will let you near the patients. You can achieve your research goals with just the PhD and you will likely be better able to achieve your research goals if you don't spend 4-10 years pursuing an entirely new domain. My research goals are pretty much on hold until I have some research time during residency, 4-5 years from now. It is highly possible the pushback you are experiencing is simply because healthcare is an incredibly slow moving obstinate beast and any change will be met with the highest level of resistance.

Do you want to have patients?

 

[longhaul3]

When you say income drop, do you mean drop from your current income to zero income as a med student or from current income to an average physician's salary, i.e. you already make 300-400k+?

IMO it would be crazy to leave a large six-figure salary with a terminal degree in a burgeoning field to go to med school. You're talking 10 years with zero to negligible income to make less than you started with AND take call.

If you just want to go for the MD then go back to industry and not practice, you would be putting yourself through a lot of pain to do a job you're already getting paid for.

 

[PubPhD]

The only reason to get the MD is if you want to personally treat patients. Without an MD, no one will let you near the patients. You can achieve your research goals with just the PhD and you will likely be better able to achieve your research goals if you don't spend 4-10 years pursuing an entirely new domain. My research goals are pretty much on hold until I have some research time during residency, 4-5 years from now. It is highly possible the pushback you are experiencing is simply because healthcare is an incredibly slow moving obstinate beast and any change will be met with the highest level of resistance.

Do you want to have patients?

I appreciate your input. Talking with some current MD students made it seem like research represents a significant portion of their time spent in medical school and you don't need to interact with patients that much (unless you decide you want to). This seemed drastically different when I was looking into med schools way back in undergrad, even with the new curriculum changes that have been occurring, so I wanted to get a larger sample. It's also good to know that I'd still receive this pushback even from the other side!

When you say income drop, do you mean drop from your current income to zero income as a med student or from current income to an average physician's salary, i.e. you already make 300-400k+?

IMO it would be crazy to leave a large six-figure salary with a terminal degree in a burgeoning field to go to med school. You're talking 10 years with zero to negligible income to make less than you started with AND take call.

If you just want to go for the MD then go back to industry and not practice, you would be putting yourself through a lot of pain to do a job you're already getting paid for.

It would be both, but I figured that I would still make enough after med school to not worry about money in the long term. The idea would be to stay in healthcare rather than go back to industry.

 

[Naruhodo]

Honestly it depends what you want out of life (and no one but you can answer this question). It is however worthwhile to try to gather as much info as you can to input into your decision-making process. I tend to agree with @amuseddoughnut above that whether you want to see patients or not is a good (if not entirely comprehensive) litmus test.

Are there medical doctors that do important work and don't see patients? Absolutely, but sometimes these jobs are ones that don't strictly require the MD (even if it they may be easier to get into with it). Did you know that there are 2 medical specialties whose boards certify PhDs as well as MDs? As you may have guessed, they are specialities that don't necessarily require patient-contact (radiology and laboratory genetics/genomics). If your vision a future career doesn't involve patient-contact, see if you can figure out a way to do the work you want to do without an MD (machine learning is already impacting analysis of radiological images, diagnosis of rare diseases with distinctive facial characteristics, as well as how electronic health records are built/analyzed). Will you have to jump through hoops that MDs don't necessarily have to? Quite possibly, but if you save 7+ years of additional school/training it may be worth it.

If you do want to see patients, that is a different kettle of fish. I've met a couple physicians who do work similar to what it sounds like you may want to do (one works full time for google these days), but even they spent years participating in the hands-on care of patients in the hospital and still stumble upon roadblocks when trying to access information (you can thank HIPAA as much as the insularity of physicians for that). And don't assume a lucrative career will necessarily materialize just because you are an MD (it may, but just don't base your decision on it). For me, the job I'm most interested in will realistically take an additional 9 years of education/training and the starting salary will most likely be $20k/yr less than what I'd make otherwise. I've made my peace with that, but it's a personal decision.

 

[ellealla]

I think you should do an MD. I am doing an MD and a PhD in computer science (specifically artificial intelligence) and having an MD makes a HUGE, HUGE, HUGE difference! Benefits: (a) doctors will trust you if you have an MD; (b) you will speak the language of medicine; (c) most importantly, you will have a great perspective on what kinds of problems are important, what kinds of problems are out there, and how to solve them in a way that makes sense. Just the other day I talked to someone building a machine learning algorithm to interpret a certain kind of radiological scan, and they were building the model in a completely bizarre way because they had no idea how the scans were read, and what the "Bad Thing" in the scan actually looked like. I talked to them for an hour about how the scan actually works and how you interpret the scan, and we came up with a completely new way of buliding the model that is guaranteed to work way better than what they were doing before (Which would have led to some really weird, medically implausible assumptions.) If you do not want to do the MD yourself, thenyou should collaborate heavily with people who are MDs. I think machine learning is going to revolutionize medicine and we need people who understand both sides. So I definitely encourage you to get an MD even if you don't want to do residency. it will change your perspective and help out a lot with your research. feel free to message me if you have any other questions about combining artificial intelligence and medicine. good luck!!! 🙂

 

[trev5150]

I don't think the pushback you're experiencing is because you don't have an M.D. It's probably more likely that you're trying to introduce another level of computer control and second-guessing on docs that already have enough oversight and second guessing by admins, insurance, patients and their families. Inserting yet another computer into their already complicated lives that would take away from meaningful patient contact and leveraging years of study and experience with and endless ribbon of "Are You Sure?" prompts from an immature technology is like asking them if they want to voluntarily attempt to suck-start a revolver.

Not saying that's your goal, but that's probably the perception from the docs. The med students you've talked to don't have the full picture. You need to understand what's up out in the street, in the hospitals from big to small.

Right now all anyone sees AI for is a toy, a joke, because if it was so great, then Siri and Cortana and Alexa wouldn't be the annoying stodgy junk they are, and that's the state of the art from Google, Apple and Microsoft - representing the biggest R&D budgets on the planet.

If you really believe in the thing, do your post-doc and develop your ideas into a usable product that is a true force multiplier that doesn't add work, hours and and erode the physicians already tenuous hold on what little time they get with their patients and, ultimately, their whole raison d-être and you'll find that all concerned will the better for it.

You don't need an M.D. to make AI work in medicine, you just need to make AI not suck first for the world in general (softening the entrance of AI into healthcare) and then adapt it to medicine in a way that the medical community wants and needs.

 

[ellealla]

I think trev5150's response exactly illustrates why having an MD would benefit your application of AI to medicine. A lot of doctors don't like AI because they don't understand it and they don't want any new "tools" making their lives more difficult. Some AI experts come up with "solutions" that are annoying because they don't address a real need. But, many AI experts (those with medical background or at least medical collaborators) come up with extremely useful AI-based tools that save doctors time, or assist them in some other way. For example, I know of a huge group of radiologists who are collaborating with AI experts to build a system that will prioritize scans for them to read. Note: this isn't "replacing the radiologists". It's simply re-ordering the scans so that instead of seeing scans in a completely random order, the scans with more serious pathology are read earlier, so that more urgent patient decisions can be made sooner. AI can also be used to predict medical outcomes so that patients who are highly likely to have a bad outcome in the future can receive more preventative care. The whole point of AI in medicine should be to streamline, so that doctors can spend MORE time with patients, and patients can receive better care. Also, the fact that Siri and Cortana and Alexa can understand even half of the words you speak, translate them to text, and know what that text means in order to do something (like pull up directions to a location) is impressive in my book. For that matter, a lot of maps applications that give you directions use AI themselves. Google search results use AI; Amazon's predictions of what you want to buy next use AI; weather reports use AI; all self-driving cars use AI; all speech processing, text processing, face recognition, financial predictions, image captioning, etcetera are built on AI. So, I stand by my previous recommendation...AI is going to make medicine better, whether doctors like it or not, and it's going to happen through strong collaborations between doctors and AI experts, and AI experts who understand medicine themselves.

 

[PubPhD]

.