Change my mind: Medicine can be automated in 10 years

[siliso]

AI can’t even read EKGs as mentioned. Or even apparently identify all the storefronts in a picture. It’s going to learn subtle physical and emotional signs, nuanced complex communications, and shared decision making in settings of massive uncertainty and conflicting data? Not soon, don’t think so. Maybe after the singularity.

---

Members don't see this ad.

 

[libertyyne]

I think you are overestimating the technology underlying driving ai. Just to clarify what is going on:
- object detection. This is the first endeavor in AI that really took off and now anyone with a RPI2 can have a camera that draws boxes around different objects and name them in close to real time. Self-driving cars further optimize already created models based on the new data they constantly collect and their predictions get filtered/optimized for pedestrians, bikes, cars, curbs, lane markings etc.
- prediction of velocity of objects/their future position (is this bike static or moving into the car’s path)
- integration of other data that a human may not pick up on (weather, wind, prediction of a car 2 cars ahead that the driver can’t even see etc.
- +/- hierarchical rules like following street signs, lights etc.

It’s not that these things aren’t incredibly complex in practice and at the front end, but that the underlying tech is not that crazy at all. Self-driving cars are one of the first consumer level applications for
1) simplified predictions needed (object classifier and velocity/position which is vector/matrix math that computers excel at)
2) ease of data creation. Any car with sensors literally generates perfect data 24/7 while driving
3) limited outputs based on predictions (steering, gas, break, lights)
4) immediately improves on the current situation even when unoptomized


Medicine has both 1+2+3 in very few fields with the biggest issue with 2. In fact to solve 2, there are AI based deidentification projects going on right now to scrub PPI from nOtes using NLP which could be a HUGE win for the future of AI in predictive aspects of medicine.


In the computer vision world, classifier data is rectangular bounding boxes around labeled objects in images. Im hoping for the day an AI could reliably trace the outline of perceived significant findings and landmarks and spit out calculated sizes, delta from previous studies etc in real time US, CXR etc. before focusing on dxing pneumonia or cancer

I think you you are underestimating the complexity of each of those pieces. Pedestrian identification, and road sign identification by itself is a huge undertaking. Even lane detection during road work, or poorly marked lanes requires quite a bit of detection complexity. Plus lets not forget all the camera, lidar, sensor tech that needs to be integrated on top. Its taken alphabet with unlimited resources and know-how close to a decade and a half to get this far.

The deidentification projects are definately going to be a boone to AI. there are already small datasets for images. I cant even imagine what would happen if there were huge ones available. And there are a few papers already indicating pneumonia identification better than or equivilant to radiologists from the group in stanford. I think there is a role of augmentation, but in a world where there is a large dataset you dont even need to be augmenting reads, it could just spit out a likely diffrential or read normal vs abnormal.

 

[libertyyne]

Also, no one brought this up, but what about the security threats with a system like this? Imagine if someone could hack the AI, and make it give a different result then what's actually the case. Could be used as a weapon of biowarfare

You could say that about everything including voting, banking, communication, power grids etc, traffic lights,etc. Not a huge consideration in the grand scheme of things considering the security right now for medical records isnt exactly fort Knox.

 

[Lucca]

I think you you are underestimating the complexity of each of those pieces. Pedestrian identification, and road sign identification by itself is a huge undertaking. Even lane detection during road work, or poorly marked lanes requires quite a bit of detection complexity. Plus lets not forget all the camera, lidar, sensor tech that needs to be integrated on top. Its taken alphabet with unlimited resources and know-how close to a decade and a half to get this far.

The deidentification projects are definately going to be a boone to AI. there are already small datasets for images. I cant even imagine what would happen if there were huge ones available. And there are a few papers already indicating pneumonia identification better than or equivilant to radiologists from the group in stanford. I think there is a role of augmentation, but in a world where there is a large dataset you dont even need to be augmenting reads, it could just spit out a likely diffrential or read normal vs abnormal.

About the Lab | Laboratory of Quantitative Imaging and Artificial Intelligence (QIAI)

rubinlab.stanford.edu

this is probably the group you’re referring to if other ppl are interested in the real life work of AI in medicine and not the fantasy-land apocalyptic stuff u read about on Reddit.

It’s also small thinking to limit yourself to what we can see with the naked eye. With a few relatively cheap adjustments a computer can see much more than that. For example, a shortwave IR probe fitted to an endoscope can significantly increase the kind of info available to an AI. Moreover, that info can be even more information dense than simply identifying features on a flat image because IR/Raman spectra also contain chemical info which people have already studied for decades to identify disease states or progression in tissue.

Another oft undiscussed Avenue is how imaging AI will merge with genomics/multi-omics oriented AI. There’s a group I interviewed with this past year which used cardiac ultrasound information and AI to find novel genetic pathways in cardiac disease and small-scale deformations. Very much in its infancy, but this will be a boon for research. Again, probably won’t replace docs anytime soon but will definitely impact the Arsenal at the disposal of biomedical science

 

[libertyyne]

About the Lab | Laboratory of Quantitative Imaging and Artificial Intelligence (QIAI)

rubinlab.stanford.edu

this is probably the group you’re referring to if other ppl are interested in the real life work of AI in medicine and not the fantasy-land apocalyptic stuff u read about on Reddit.

It’s also small thinking to limit yourself to what we can see with the naked eye. With a few relatively cheap adjustments a computer can see much more than that. For example, a shortwave IR probe fitted to an endoscope can significantly increase the kind of info available to an AI. Moreover, that info can be even more information dense than simply identifying features on a flat image because IR/Raman spectra also contain chemical info which people have already studied for decades to identify disease states or progression in tissue.

Plus the beauty of AI is that you dont need to know how the sausage is made by the trained model. As you mentioned even with imaging data, you could rely upon raw data and build models from there where it doesnt even go through an image processor for CT or MRI.

 

[Lucca]

One last thing just literally recalling the interview trail. You can gain a lot of info from what you are able to find in urine. Imagine a world where you have a smart toilet fitted with something like a very cheap, AI driven Mass-Spec which is able to warn you if anything seems abnormal. Think Apple Watch EKG with a thousand times more resolution. Also already happening and it’s the military that’s testing it out first.

When wearables and integrated biotech become cheaper and more widely adopted I suspect they will actually increase utilization of healthcare resources and increase, not decrease, demand of professionals who can work with that info in a meaningful way to help people monitor their health. Who knows, maybe primary care will be the most popular specialty for tech nerds.

 

[deleted966164]

It’d be interesting to make one of the AIs take the MCAT...

 

[zanzizic]

I don't think there will be a replacement in 10 years, but the potential good that can come of AI would be wonderful. Imagine scanning through every PubMed article and medical text book ever published in a second. I think that AI can help physicians make percentile diagnoses based upon the data that they enter during a patient assessment. For example, they enter all of the data and laboratory values from the patient encounter and then a list of possible diagnoses is generated. The physician then combs through the list and enters additional differential information to get an accurate diagnosis. If it is the wrong diagnosis, the AI has the ability to correct itself. Also, all of the computerized data can be linked on a network and improve its accuracy based on outcomes. What if this technology merged with an advanced DaVinci robot?

Since you’re a pre-med student this sounds like an amazing idea. When you become a physician you’ll realize how difficult, murky, and subjective medicine really is.

History is very important. Only humans can give their history and that’s why is so confoundingly subjective. And only doctors (not a computer) can take a history which is putting that information through another subjective filter. Patients lie all the time. The exaggerate symptoms or they minimize symptoms. An experienced physician can often (but not always) get to the truth of the matter. But even when he does....how is that physician going to put that historical narrative into a computer in a meaningful way? How do we quantify symptoms so they are compatible with symptoms that another physician is putting into the computer?

Physical examination is crucial. A computer or robot is never going to be able to perform a physical examination. Can anyone imagine how that could ever even possibly work? Ok, maybe if there is a suspicious skin lesion a specific scan could be deployed that may rule in or out malignancy. But how would a computer/robot ever look at a patient and get a sense of overall health? Check under the tongue for signs of subtle jaundice? Palpate the liver for hepatomegaly? Or pick up the faint smell of cigarette smoke on the patients clothes after he’s already lied about not being a smoker?

Ok, now we are at the real data! Laboratory findings! Surely we have some objective findings that a computer could really use, right? Well, not so fast. There’s plenty of times that a doc will look at a lab report and see a lab result that is an outlier. Sometimes it’s meaningful and sometimes it’s spurious. You see a glucose of 800 in a healthy thin marathon runner and you’re going to say that’s an error, let’s run those labs again...you see it in the sick dehydrated guy in the ER and you know he’s in diabetic ketoacidosis. This happens to me in dermatology all the time. I do a biopsy of a skin lesion that looks like a deep fungal infection and the dermatopathologist calls me and says hey this looks like squamous cell carcinoma. Well, as dermatologist who has studied pathology I know that a deep fungal infection can exhibit pseudoepitheliomatous hyperplasia that can simulate squamous cell carcinoma. My clinical correlation tells me that cancer is a red herring (clinically that did not look like cancer) so I tell the pathologist to cut further into the specimen and he ends up seeing the infectious organisms. The pure, rote pattern recognition of a computer reading pathology would have a hard (actually impossible) time making that call...

I won’t even discuss the issues that could possibly arise from an AI perusing the medical literature. How would it know which papers are of historical interest only? I know the surgeon personally who wrote those papers on surgical margins was on a crusade and frankly full of **** and nobody really believes that...how is the AI going to have that context?

As someone said before...medicine is a human endeavor. It’s truly an art, not a science. Yes we wield science as a tool, but you cannot quantify the art and put it into a computer and make that work. It just won’t.

So frankly, a DaVinci robot armed with PubMed and dubious input of symptoms and physical examination (again, how do you quantify that stuff?) and devoid of human experience and intuition actually sounds like a dystopian nightmare.

 

[LoGo]

I could see a near future where NLP is good enough to significantly improve efficiency/documentation in the clinic. Anyone seeing the pt wears a mic. It’s turned on during pt encounter. Human asks questions / makes observations. Mic records. Robot takes that audio data, parses for keywords and generates the note. Backend AI assists with using that data to bring up a probabilistic differential. Doc makes edits as needed/corrects any errors and uses AI info to make final decision on what to do.

I also think AI integrated with imaging technology will assist rads and path in reading slides/images. That’s already happening. In the farther future AI will be integrated with live imaging from scopes and be able to assist in pointing out features of interest on video reducing human error during procedures or in deciding treatment. This is technology that is already being developed in labs so I’m fairly confident it’ll exist in some commercial form by the time my generation is attending.

I don’t think any of it will present an existential threat to physicians but it will change the way people train and practice.

I do foresee this tech creating new ethical and organizational challenges for healthcare, for example by necessitating the creation of bodies or other technology capable of monitoring the use of AI assistants to ensure they aren’t being gamed to over-Bill or over-use resources

If they still teach how to percuss a lung in med school in a patient with cough and fever when you know an x-Ray is right there, all of the space age stuff you just mentioned sadly will not happen for I estimate 300 years.

 

[libertyyne]

If they still teach how to percuss a lung in med school in a patient with cough and fever when you know an x-Ray is right there, all of the space age stuff you just mentioned sadly will not happen for I estimate 300 years.

like the physical exam is known for its high inter-rater reliability and and sensitivity and specificity.

 

[LoGo]

like the physical exam is known for its high inter-rater reliability and and sensitivity and specificity.

Sorry, it sure if you mean this sarcastically or not.

In any case, my point is just that adoption of new approaches in medicine and getting rid of ancient ways that aren’t particularly useful on their own is very slow in medicine, regardless of what all the latest hot research and media would have us think.

 

[Dr. Brightside]

How come the IT field isn't automated yet, even though it's significantly more formulaic and requires less creativity and human interaction than medicine? Or why not literally any field outside of manufacturing?

 

[libertyyne]

How come the IT field isn't automated yet, even though it's significantly more formulaic and requires less creativity and human interaction than medicine? Or why not literally any field outside of manufacturing?

Phone tree's
automated web traffic balancing /automated server recruitment
There are IT departments with access granting automated etc.

Also IT labor is cheaper compared to the cost of automating it. Medicine is not. Its cheaper to purchase support staff in india compared to automating it.

 

[KeikoTanaka]

This website is wild. I’d say less than 5% of “for sure predictions in under 10 years” that happen on here ever actually come true. If they did, every specialty would be getting paid pennies and robots and midlevels would rule the world

Or midlevels would phase out because why would you want a Midlevel and a Robot when you can have a doctor and a robot? I'd rather have a Car Mechanic and a Robot fixing my Car than the car Mechanics assistant and a robot fixing my car. Now if the robot is doing something weird, what if he doesn't know how to correct it? If anything the midlevel tasks can be taken over by the robot, not the physician.

Phone tree's
automated web traffic balancing /automated server recruitment
There are IT departments with access granting automated etc.

Also IT labor is cheaper compared to the cost of automating it. Medicine is not. Its cheaper to purchase support staff in india compared to automating it.

Also, some IT problems are HARDWARE, not software, sometimes you just need a human touch to fix something. Like a good spinning-eagle kick to the tower to make it boot up again.

 

[Dr. Faust]

Since you’re a pre-med student this sounds like an amazing idea. When you become a physician you’ll realize how difficult, murky, and subjective medicine really is.

History is very important. Only humans can give their history and that’s why is so confoundingly subjective. And only doctors (not a computer) can take a history which is putting that information through another subjective filter. Patients lie all the time. The exaggerate symptoms or they minimize symptoms. An experienced physician can often (but not always) get to the truth of the matter. But even when he does....how is that physician going to put that historical narrative into a computer in a meaningful way? How do we quantify symptoms so they are compatible with symptoms that another physician is putting into the computer?

Physical examination is crucial. A computer or robot is never going to be able to perform a physical examination. Can anyone imagine how that could ever even possibly work? Ok, maybe if there is a suspicious skin lesion a specific scan could be deployed that may rule in or out malignancy. But how would a computer/robot ever look at a patient and get a sense of overall health? Check under the tongue for signs of subtle jaundice? Palpate the liver for hepatomegaly? Or pick up the faint smell of cigarette smoke on the patients clothes after he’s already lied about not being a smoker?

Ok, now we are at the real data! Laboratory findings! Surely we have some objective findings that a computer could really use, right? Well, not so fast. There’s plenty of times that a doc will look at a lab report and see a lab result that is an outlier. Sometimes it’s meaningful and sometimes it’s spurious. You see a glucose of 800 in a healthy thin marathon runner and you’re going to say that’s an error, let’s run those labs again...you see it in the sick dehydrated guy in the ER and you know he’s in diabetic ketoacidosis. This happens to me in dermatology all the time. I do a biopsy of a skin lesion that looks like a deep fungal infection and the dermatopathologist calls me and says hey this looks like squamous cell carcinoma. Well, as dermatologist who has studied pathology I know that a deep fungal infection can exhibit pseudoepitheliomatous hyperplasia that can simulate squamous cell carcinoma. My clinical correlation tells me that cancer is a red herring (clinically that did not look like cancer) so I tell the pathologist to cut further into the specimen and he ends up seeing the infectious organisms. The pure, rote pattern recognition of a computer reading pathology would have a hard (actually impossible) time making that call...

I won’t even discuss the issues that could possibly arise from an AI perusing the medical literature. How would it know which papers are of historical interest only? I know the surgeon personally who wrote those papers on surgical margins was on a crusade and frankly full of **** and nobody really believes that...how is the AI going to have that context?

As someone said before...medicine is a human endeavor. It’s truly an art, not a science. Yes we wield science as a tool, but you cannot quantify the art and put it into a computer and make that work. It just won’t.

So frankly, a DaVinci robot armed with PubMed and dubious input of symptoms and physical examination (again, how do you quantify that stuff?) and devoid of human experience and intuition actually sounds like a dystopian nightmare.

Medicine is becoming increasingly dependent upon machines and computers and there will be more and more AI involved as time goes on. Dystopian nightmares have predicted things in medicine such as in vitro fertilization (Brave New World). Laboratory tests, MRIs, digital x ray enhancements, etc. all use science more than art. Physicians utilize cell phones more than fancy quill written calligraphy notes. Statistics can be applied to medicine and therefore refined by AI. Art is completely subjective as one cannot prove what is beautiful. "de gustibus non disputandum est,"You an prove which treatments are more successful. Therefore the "more of an art than a science" argument is more fit for a theologian than a physician.

 

[premeddestiny]

Medicine is becoming increasingly dependent upon machines and computers and there will be more and more AI involved as time goes on. Dystopian nightmares have predicted things in medicine such as in vitro fertilization (Brave New World). Laboratory tests, MRIs, digital x ray enhancements, etc. all use science more than art. Physicians utilize cell phones more than fancy quill written calligraphy notes. Statistics can be applied to medicine and therefore refined by AI. Art is completely subjective as one cannot prove what is beautiful. "de gustibus non disputandum est,"You an prove which treatments are more successful. Therefore the "more of an art than a science" argument is more fit for a theologian than a physician.

Straw man argument. These scientific advancements (IVF, MRI) are all tools used to facilitate the art of caring for patients unlike automated robots like a robot obtaining history/physical from a patient.

 

[KeikoTanaka]

A Doctor with an AI machine assisting them will be much better than an AI machine alone. #ThanksForComingToMyTedTalk

 

[Dr. Faust]

Straw man argument. These scientific advancements (IVF, MRI) are all tools used to facilitate the art of caring for patients unlike automated robots like a robot obtaining history/physical from a patient.

My point is demonstrating how medicine is more of a science than an art related to all of the science involved. How many hours do physicians spend studying art? How many hours do they spend studying physiology, biology, anatomy, and biochemistry? Also humans are notoriously bad at detecting truth vs. lie. No straw man here, glad you took a 100 level logic or philosophy class though.

 

[Dr. Faust]

A Doctor with an AI machine assisting them will be much better than an AI machine alone. #ThanksForComingToMyTedTalk

Will a doctor with AI be better than a doctor without AI?

 

[Thesimplelifeofamyloid]

My point is demonstrating how medicine is more of a science than an art related to all of the science involved. How many hours do physicians spend studying art? How many hours do they spend studying physiology, biology, anatomy, and biochemistry? Also humans are notoriously bad at detecting truth vs. lie. No straw man here, glad you took a 100 level logic or philosophy class though.

Dude no offence, but you're premed. Medical school doesn't equal medical practice, so undergrad sure as hell doesn't equal anything. Even I know that, and I'm the one that created the thread in the first place

 

[ortnakas]

Admittedly, didn’t read much of the thread. But the “robots” in EKG machines only do their job right 50% of the time, and that’s electricity reading about electricity. I think radiologists and even moreso the rest of us are safe for a while.

 

[Dr. Faust]

Dude no offence, but you're premed. Medical school doesn't equal medical practice, so undergrad sure as hell doesn't equal anything. Even I know that, and I'm the one that created the thread in the first place

No offense taken, but did physicians predict the current ailments of the healthcare system in America? Did physicians predict that NPs would function in a nearly independent state? Being good at medicine and predicting the future are two completely different domains. This follows the fallacy of "I'm a doctor, ergo I can do anything."

 

[Lawpy]

What’s with SDN’s obsession with AI? These threads keep popping up regularly with similar arguments discussed

 

[Dr. Faust]

What’s with SDN’s obsession with AI? These threads keep popping up regularly with similar arguments discussed

Fear of something greater than themselves. Gods don't like competition.

 

[Sephirakra]

Medicine is becoming increasingly dependent upon machines and computers and there will be more and more AI involved as time goes on. Dystopian nightmares have predicted things in medicine such as in vitro fertilization (Brave New World). Laboratory tests, MRIs, digital x ray enhancements, etc. all use science more than art. Physicians utilize cell phones more than fancy quill written calligraphy notes. Statistics can be applied to medicine and therefore refined by AI. Art is completely subjective as one cannot prove what is beautiful. "de gustibus non disputandum est,"You an prove which treatments are more successful. Therefore the "more of an art than a science" argument is more fit for a theologian than a physician.

My point is demonstrating how medicine is more of a science than an art related to all of the science involved. How many hours do physicians spend studying art? How many hours do they spend studying physiology, biology, anatomy, and biochemistry? Also humans are notoriously bad at detecting truth vs. lie. No straw man here, glad you took a 100 level logic or philosophy class though.

This just in: the term "art" has more than one definition. Film at 11:00.

 

[deleted819639]

Some great discussion here, but I think people here are pretty quick to dismiss AI's capabilities. Sure, right now AI is limited to simple image processing, screening, and NLP tasks, and recent demonstrations have not been too impressive. (Although I'm not sure why people are mocking AI's capability to read EKG, since a 2017 paper from Andrew Ng at Stanford had a convnet that outperforms individual cardiologists in diagnosing arrhythmias from ECGs!) However, AI is a rapidly developing field, and IMHO AI applications in medicine are lagging far behind what is possible at the frontiers of AI research. Much of the state-of-the-art NLP, GP programming, etc. have not yet seen wide-spread implementation in the clinics, preventing us from making an accurate assessment as to how useful they would be in medicine. A lot of that has been due to physician resistance and skepticism that are quite apparent in this thread.

I absolutely agree with others that AI will never replace the "human" aspects of diagnosing and treating a patient, and I don't see AI ever taking away jobs from physicians. But the decision-making steps involved in a diagnosis are (or should be!) consistent, methodical, and rule-based, and there is nothing stopping AI from learning them. In the future, it is possible that one of the jobs of a physician will be to procure accurate data from the patient, plug it into a machine, and then verify and relay the output to the patient.

 

[Dr. Faust]

This just in: the term "art" has more than one definition. Film at 11:00.

Art is undefinable. Are you familiar with the Dadaist movement or are you still gushing over your starry night shower curtain?

 

[Sephirakra]

Art is undefinable. Are you familiar with the Dadaist movement or are you still gushing over your starry night shower curtain?

Everyone is beneath your intelligence, right? Reported and ignored.

 

[Dr. Faust]

Everyone is beneath your intelligence, right? Reported and ignored.

I didn't say anyone was beneath my intelligence. So you just report people who disagree with you? You'll be a great one day

 

[Thesimplelifeofamyloid]

No offense taken, but did physicians predict the current ailments of the healthcare system in America? Did physicians predict that NPs would function in a nearly independent state? Being good at medicine and predicting the future are two completely different domains. This follows the fallacy of "I'm a doctor, ergo I can do anything."

Yeah but neither did a premed. People at your level tend to think they know more then they actually do

 

[GoSpursGo]

This doesn't really have anything to do with medical school--moving to Topics in Healthcare.

Also, please refrain from personal attacks.

 

[KeikoTanaka]

Will a doctor with AI be better than a doctor without AI?

Of course. That's a dumb question, no offence, that's like asking if an Orthopedist with an MRI machine is better than an Orthopedic without an MRI machine. OF COURSE the Orthopedist with an MRI will be better than the one without as you can get a picture of what's going on before you aimlessly cut into someone.

Obviously technology augments decision making processes and makes life easier for Physicians. Burnout is so high right now because the technology moved faster than the older generations of physicians could keep up. I scribed on the "Horrible EMRs" that hospitals complain about daily - They complain because 54 year old Susan has no idea how to recall patient records faster than looking through a filing cabinet, and they can't type more than 20 words / minute, stressing everyone out. Why was I not stressed out? I've been playing online games since I was 8 years old and can type 120 wpm, with the right amount of coffee and focus.

That's why this whole "Heart of a Nurse, Brain of a Doctor" thing has been being pushed by NPs imo - Older Drs can NEVER type and maintain eye contact with their patient - making them look like this "Heartless cold dingus" when in reality its just an old guy trying to figure out how the hell to pull up old labs on a new EMR and can barely type 20 WPM to keep up with the patient's history. So now the Physician is behind on schedule when he has administration breathing down his neck to see more and more patients so not only did they barely look at you - but they hurried through the appointment and didn't examine you. Because the technology that was supposed to make their lives easier was in fact a barrier for them.

Technology isn't going to replace physicians, it's going to enhance the work a single physician can do while making the profession safer and more efficient. Older physicians will be weary of it, newer physicians will embrace it, like all technology in all walks of life/professions. Eventually I'm sure there will be Virtual Reality Physicians controlling a robot from elsewhere making calls while wheeling around a Hospital making RNs administer everything. All the while an AI camera in every room will be feeding the physician pertinent information and filtering out anything that they don't need to see/hear. But ultimately, the AI will never make the FINAL call. And there will always be a need to have a human life present in the event there is a power outage or the inability for the robot to do a tough procedure, like a difficult airway, or palpating an abdomen, or performing a fragile rectal/vaginal exam. There's just some things that can never be replaced by robots.

 

[zanzizic]

AI can make things even more inefficient, too. Now that I have EMR, every time I try to prescribe a drug the program throws up one or two “interactions” with the patients current meds and the MAs start freaking out. I have to explain that “No, there’s no problem with Augmentin and lasix together...just override the EMR and send the script”. Who knows why there is that alert? Maybe there was a case report of an interaction?

Now I have to override these all the time and I’ve never had an issue.

Of course, EMR isn’t AI, but garbage in/garbage out. Sometimes what people program to “help” the physician actually makes is slower and more inefficient.

 

[Mad Jack]

Computers are only as good as the inputs provided, and only a competent physician is good enough to provide the inputs required for these algorithms to function properly, so no, they won't be replacing physicians any time soon.

 

[GH253]

Even if medicine could be automated with indisputably better results, we’d never realize that benefit, because physicians would lobby to convince legislators that robots are the greatest threat to The Public Health since thalidomide.

 

[Sephirakra]

Even if medicine could be automated with indisputably better results, we’d never realize that benefit, because physicians would lobby to convince legislators that robots are the greatest threat to The Public Health since thalidomide.

How's that tinfoil hat fitting these days?

 

[GH253]

How's that tinfoil hat fitting these days?

How’s that delusional fantasy world you live in where physicians serve as Jedi guardians of inferior beings’ welfare or do anything other than protect their own turf?

 

[Sephirakra]

How’s that delusional fantasy world you live in where physicians serve as Jedi guardians of inferior beings’ welfare or do anything other than protect their own turf?

That escalated quickly, and predictably. Aaaaaaaaaaaaand ignored.

 

[SurfingDoctor]

Yet, somehow computer models can only predict a 20% chance of rain.
Yeah, no AI will not take over medicine in 10 years.

Maybe someday, but not in the near future. What you may see is integration of AI into EMR to give physicians alerts or advice on treatments... until physicians get so annoyed will all the alerts, they get fed up with the computer and throw it out the window.

 

[ArdorAyurveda]

Would be hilarious to have a robot measure the height and weight of a bodybuilder, then classify him/her as "obese"

Or even better, performing a mental health exam on literally any human being.

lol