Diagnosis is based on imperfect data with very high noise-to-signal ratio, and with auditory (patient's history), visual and tactile inputs. Treatment often need to be tailored for each patients unique needs, goals, and co-existing diseases.
But I agree that the future is likely to change, as it becomes easier for non-medical persons to access and understand medical information and recommendations. It will just then come down to regulatory barriers. When will a computer be given a medical license? 20 years? 50 yrs? 5 yrs?
Often what "AI" is doing is what a doctor with good active listening skills has, but until the field gets more diverse (and the ones who make it in actually use some of their humanities lessons rather than treat being an MD as a chance to act like a rude old white man) nothing substantial will change beyond more 1920s style civil unrest, which occured around the time antibiotics where discovered and germ theory became accepted.
Also, rich people and politicians, will live longer as usual. Expect Elon news for another century!
Instead of using ML to diagnose, we are treating the student as the AI. They're given chatbots which they must treat and diagnose. Fortunately for us clinical sessions usually follow a formula, and so make for a good case for chatbots. (We're called SimConverse in case anyone is interested)
Not in common use, and doctor's role hasn't changed.
AI/ML might be used for targeted medicines, prevention or early detection which will free up time for doctors.
I don’t see ML have much, if any, impact on a doctors role in diagnosis. I know this is a little strange to hear to a lot of techies, but we once had IBMs Watson work on our data to see what IBM and Watson could come up with, and while the output wasn’t bad, it was sort of useless because we already had 50 years with of analytical models on virtually everything. So in essence what Watson did with our data, was to generate a lot of BI that was inferior to the BI we already had. Diagnosis is sort of the same story because what doctors do is basically to refer your symptoms directly with the medical lexicon and your history. ML is likely going to help collect the data and tie it together, to make diagnosis both better and faster, but you’re likely still going to have a doctor review the results, just like you would your “data scientists” on any other important ML data. Where ML will change things is in early detection, where your medical data from every source will be crossed and checked much better because that’s the sort of things ML is good at. So maybe you’re going in for a blood test for something completely unrelated to cancer, but because the tech has improved, it’ll also screen you for cancer and alert your doctor if your numbers are off.
Where we’ll see the biggest impact will be in areas that can be turned robotic. Things like the lab work on your blood tests, which is currently still a very manual process. Or in surgery where precision machinery will slowly take over a lot of the cutting. You’re likely still going to need surgeons to monitor the process, not so much the actual surgery but the planning and the recovery, because these things are impacted by so many real world factors that we may never get ML models that are good enough to handle them on their own.
It’s often in the places you may least expect it that IT makes the biggest difference. The biggest impact I saw in the medical system was automated medicine distribution and in wound cleaning. The medicine distribution “used” to be handled by nurses putting the pills for patients into boxes and then healthcare workers administering it the right time. The automated way was having the pharmacy distribute smart-dispensers that would alert citizens to take their medicines and then sort of “punch” the right pills into a tray when the patient clicked on it. I put “used” in “ because the automated smart system is more expensive than the old way, and this resulted in many places still opting for the old way, despite it being more error prone. The wound cleaning is an AR/ML success story. Basically wounds can be really nasty, and contain nasty things that even the best nurses in our system won’t spot as well as ML. So what we did was hack a pair of Google glass to never send data to Google (I believe Google was very helpful in this process by the way) but instead feed images of the wound to ML recognition and then alert the nurse to areas in the wound where the nurse had missed a spot of nasty. Really awesome stuff.
In general I expect that medical, along with farming, tech will be some of the most interesting tech areas in the next few decades, but I don’t expect either to replace doctors or nurses. I expect to see it make doctors and nurses better at their jobs. It will free up some tasks, and change which doctor professions get the highest pay because a neurosurgeon won’t be the Hollywood rockstar, but really, they kind of aren’t outside of Hollywood anyway, at least no in Denmark.
As paviva mentions diagnosis is based on imperfect data with a high noise-to-signal ratio. Which I think is actually a testament to human's ability to navigate all of that and still be able to care and treat illness.
I think the trends in technology with healthcare will more likely be to support and supplement humans rather than taking them out of the equation. I do agree that there will probably be a shift in how we practice as in how much we need to retain as doctors. Doctors in the next 5-10 years will certainly need a much greater understanding how to critique ML used to make recommendations. I doubt clinicians will need to understand the inner depths of how an ML model is working at a code level though. There are already several guidance papers on how to evaluate ML models.
But clinical reasoning will still be a human endeavour in my opinion. The adjuncts to diagnosis is already happening in more visual concentrated specialities like radiology,pathoogy and ophthalmology, where ML algorithms I think have a good fit. No doubt this is where we will start to see more reliance on the ML models. But again, that uncertainty and ability to take an imperfect data set and make a decision with conscious and subconscious understanding will be aided by these models but not necessarily determined by them. To be able to capture all of the other inputs that are required for diagnostic reasoning will still escape ML inspired technology. Blood tests and imaging alone do not make a diagnosis, its the whole context from symptom to investigations together.
Another area of technology is the use of hand held devices for diagnostics, portable US for example. US is known to have better accuracy than chest X-rays for particular conditions. It's not quite there yet in terms of the taking that on a ward round, but that feels like a more realistic proposition in the next 5-10 years.
I also expect that we will be looking far far more about patterns in this noise of data. Looking at 100,000's of medical records to try to understand the patterns of disease we as humans just can't perceive. Looking at predictive modelling for things like cancer. Cancer being a large bucket of atleast 200 types of diseases, most of them are rare and no one clinician has seen enough of them to really get predictive or understand the nuance of them to really diagnose with confidence. Additionally clinical trials are poor datasets for sample sizing and in application to real life scenarios that we end up having to rely on.
So being able to get through millions of notes and finding all these cases to make predictions seems to me to be a fruitful area of investigation.
