Will doctors become obsolete?

Started by pgoyal
This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.
Get help with your application

Use all the free resources available to you from SDN: articles, guides, expert advising, forums discussions, and school research.

Browse articles Compare schools My application plan Read Interview Feedback
Practice for interviews Ask an Advisor
Advertisement - Members don't see this ad
peteB said:
Fine. If you guys don't want me here then I'll leave, I don't want to be somewhere I'm not wanted.
Click to expand...

I actually find these threads pretty interesting and I feel that you have a lot to add being that your in med school and seem to be experienced in the comp sci/AI field. Your posts come across as excessively argumentative and defensive anytime someone posts something contrary to the idea that computers will one day take over a physicians job/ medicine can be automated/ 80% of medicine is routine etc. I'm constantly reminded of Raymond Kurzwiel when I read your posts... the futurist who's been hawking for years that the "technological singularity is near" when it fact it hasn't happened yet and probably never will. All the docs I've talked too have told me how they are reminded on a daily basis how difficult medicine is so if you really think parts of it can be automated then more power to you.


Truth is we have no idea what the future will be like... to quote a sdn user (forget which name):

"There were computers that could beat me at chess or Jeopardy! when I was 5 or 10; so what? And I'm smarter today than I was when I was 5 or 10. Am I then more sentient? Is Bobby Fisher or Ken Jennings more sentient than either of us? Are computers today "kind of" sentient? Is sentience an all-or-none, or does it vary by degrees?

The whole thing is predicated on some underwear gnome logic:

1) Computer
2) Smart computer
3) Really, really smart computer
4)....
5).....
....
X) Computer that's self-aware, want to replicate like a life form, fears death, competes with humanity, " Not happening.
 
Also realize that these computer systems are being conceived of and built by us humans... Remember Craig Venter? The scientist who's team constructed a synthetic bacterial genome that gave rise to viable colonies? What about biological engineering... I'd argue that we'll be able to engineer biological systems that rival any computer.

"A replicating cell in your body has all the same elements that a computer does: inputs and outputs, calculating of conditions, varied responses, feedback control mechanisms, signaling, etc. And? The calculating power in a replicating nucleus far exceeds the calculating power of the comupters that existed in the time when 2001: A Space Odyssey was written. Is anyone afraid that their individual cells are going to become self aware?"
 
Advertisement - Members don't see this ad
link2swim06 said:
I hate this topic. It surfaces every 1 month on SDN. Just when every on this board finishes talking about a ridiculous idea it surfaces again.
Click to expand...

Sorry, I should have prefaced the question more. perhaps I should have asked "how advanced computers/AI will be integrated into the medical profession, primarily regarding them taking over tasks normally considered physician duties?" instead.

the reason I posted this is because I've been shadowing neonatology for the past few years where from my observation about 75% of the material discussed on rounds involves looking at lab charts/xrays and following/remembering a set protocol. just seems it would be easier (and more reliable) to have a program where you have all your SOPs loaded and you insert the lab results, etc with the computer giving you a recommendation on what to do. I'm guessing some of this is true for the adult specialties as well.

personally i want to go into this field because I like working with infants (not being able to ask the patient whats wrong brings a problem solving aspect to it, etc) and ensuring the family gets its baby back asap is always rewarding when they thank you. there's some satisfaction in resolving an "emotional" situation such as complicated pregnancies, caring for a premature baby, etc

however, i;m afraid society (as >60% of neonatology patients are on some sort of government care) wont see it really necessary to pay 200k to someone whose primary "skillset" involves following complex algorithms. or at the minimum the demand for neonatologists wont be as high - you might need a couple to oversee the unit.

Now, while making a **** ton of money isn't my primary concern, i do want to know that coming out of fellowship in my late 30s i'll be able to pay back loans, afford a house, a wife, be able to travel once a year and most of all still be doing something i love. [on a side note i looked into the daycare business and it's very women dominated - i guess imagine having to get a nanny for your kids ... would you prefer a male or female nanny, etc?]

to a premed whose yet to see the various sub specialties and the broad scope of medicine (just not possible time and resource wise right now) I thought I'd turn to this forum to get some input on what I can expect this career to really look like. obviously who the **** knows whats going to happen 10 years from now, but educated guesses are still helpful.
 
Last edited:
pgoyal said:
Sorry, I should have prefaced the question more. perhaps I should have asked "how advanced computers/AI will be integrated into the medical profession, primarily regarding them taking over tasks normally considered physician duties?" instead.
Click to expand...

This is a vastly different question than "will computers take over?" I don't foresee computers being adept at using "clinical judgment" any time soon. Yes, there will be more integration of technology with medicine in the future (look at EMR, for example), but will they "take over" physician duties? I don't think so, unless it's a minor, not-super-important-for-patient-care task.

pgoyal said:
the reason I posted this is because I've been shadowing neonatology for the past few years where from my observation about 75% of the material discussed on rounds involves looking at lab charts/xrays and following/remembering a set protocol. just seems it would be easier (and more reliable) to have a program where you have all your SOPs loaded and you insert the lab results, etc with the computer giving you a recommendation on what to do. I'm guessing some of this is true for the adult specialties as well.
Click to expand...

From what I understand, it's very hard to get machines to accurately "read" images, especially considering the amount of anatomical variation in the human population. As someone mentioned earlier, even with something relatively simple, like EKGs, we haven't come up with a computer that accurately gives the diagnosis and is "better" than board-certified cardiologists. At least, none that I've heard of.

One thing I've learned in med school (and even before then, from working in academia) is that protocols/algorithms will always break down. They don't cover every scenario. And a lot of patients present atypically, rather than in a textbook manner. As a premed, I don't think you're in a position to evaluate what a board-certified physician is thinking of when discussing a protocol on rounds, what they've considered, what they've ruled out, whether they're going to strictly adhere to the protocol or make adjustments based on experience, etc. Hell, even I'm nowhere close to judging an attending's clinical decision-making.

pgoyal said:
however, i;m afraid society (as >60% of neonatology patients are on some sort of government care) wont see it really necessary to pay 200k to someone whose primary "skillset" involves following complex algorithms. or at the minimum the demand for neonatologists wont be as high - you might need a couple to oversee the unit.
Click to expand...

Again, you're a premed and are in no position to make judgments on clinical decision-making or how complex/not-complex medicine is. This is not meant as an insult. It's just you don't have the clinical experience to understand that. I'm an M2, but I wouldn't make claims about the complexity of the practice of medicine either, simply because I don't have that extensive clinical training that an attending has. I will say, though, that I've never heard a resident or attending say how "easy" medicine is. I do often hear about how they're still learning something new on a fairly regular basis even after practicing medicine for a decade (or more).

Let me ask you a question. If practicing medicine is as simple as memorizing algorithms and spitting them back out, why does it take a minimum of 7 years to train a physician?

pgoyal said:
to a premed whose yet to see the various sub specialties and the broad scope of medicine (just not possible time and resource wise right now) I thought I'd turn to this forum to get some input on what I can expect this career to really look like. obviously who the **** knows whats going to happen 10 years from now, but educated guesses are still helpful.
Click to expand...

Don't worry. Physicians won't be replaced anytime soon. It's definitely not as prestigious a field as it was a few decades ago, but you won't have to worry about job security during our life-times, IMHO. Even if Watson were deemed fit to practice medicine independently tomorrow, it will take many years of research into efficacy and safety and even many more years afterward to incorporate that technology throughout the US. I just don't see the average person being okay with being seen by machines instead of human beings. In the surveys posted on the AMA website, the majority of patients are willing to wait significantly longer to see a physician rather than a mid-level. Do you really think they'd be more willing to see a machine to plug in their symptoms into and get an answer spit out? Will they even know what symptoms are relevant and what parts of clinical history are important in piecing together what's going on?
 
Impossible, I just finished watching Star Trek 2009 and McCoy said he went to medical school before attending StarFleet Academy. Therefore, doctors will in fact be present in the future.




For real though... space medicine sounds badass
 
pgoyal said:
Sorry, I should have prefaced the question more. perhaps I should have asked "how advanced computers/AI will be integrated into the medical profession, primarily regarding them taking over tasks normally considered physician duties?" instead.

the reason I posted this is because I've been shadowing neonatology for the past few years where from my observation about 75% of the material discussed on rounds involves looking at lab charts/xrays and following/remembering a set protocol. just seems it would be easier (and more reliable) to have a program where you have all your SOPs loaded and you insert the lab results, etc with the computer giving you a recommendation on what to do. I'm guessing some of this is true for the adult specialties as well.

personally i want to go into this field because I like working with infants (not being able to ask the patient whats wrong brings a problem solving aspect to it, etc) and ensuring the family gets its baby back asap is always rewarding when they thank you. there's some satisfaction in resolving an "emotional" situation such as complicated pregnancies, caring for a premature baby, etc

however, i;m afraid society (as >60% of neonatology patients are on some sort of government care) wont see it really necessary to pay 200k to someone whose primary "skillset" involves following complex algorithms. or at the minimum the demand for neonatologists wont be as high - you might need a couple to oversee the unit.

Now, while making a **** ton of money isn't my primary concern, i do want to know that coming out of fellowship in my late 30s i'll be able to pay back loans, afford a house, a wife, be able to travel once a year and most of all still be doing something i love. [on a side note i looked into the daycare business and it's very women dominated - i guess imagine having to get a nanny for your kids ... would you prefer a male or female nanny, etc?]

to a premed whose yet to see the various sub specialties and the broad scope of medicine (just not possible time and resource wise right now) I thought I'd turn to this forum to get some input on what I can expect this career to really look like. obviously who the **** knows whats going to happen 10 years from now, but educated guesses are still helpful.
Click to expand...

What you are saying will happen in a variation to what you described.

But everyone misses one key variable...timeframe. It is not going to happen anytime in our careers or even our lifetime. EMRs were introduced over a decade ago and there are still physicians which use pen and paper. Likewise, we don't even 'trust' the EKG machines interpretation yet.

This technology will not replace us. Maybe it is something our grand-children should worry about if they want to become physicians.
 
You guys saying that medicine is impossible for AI are ridiculous. It might be a long way off but it's more than possible once you realize that the human mind is just a biological computer. Biological circuits and humanistic systems of logic can be transformed to synthetic circuits and man-made software. Our thinking, learning, irrationalities, and intuition are not impossible to recreate.
 
circulus vitios said:
You guys saying that medicine is impossible for AI are ridiculous. It might be a long way off but it's more than possible once you realize that the human mind is just a biological computer. Biological circuits and humanistic systems of logic can be transformed to synthetic circuits and man-made software. Our thinking, learning, irrationalities, and intuition are not impossible to recreate.
Click to expand...

Link me to where anyone said that it was impossible. In this case to the multiple people.

link2swim06 said:
What you are saying will happen in a variation to what you described.

But everyone misses one key variable...timeframe. It is not going to happen anytime in our careers or even our lifetime. EMRs were introduced over a decade ago and there are still physicians which use pen and paper. Likewise, we don't even 'trust' the EKG machines interpretation yet.

This technology will not replace us. Maybe it is something our grand-children should worry about if they want to become physicians.
Click to expand...

EMR has been around for 3+ decades.
 
yesno said:
As someone who (on occasion) beat chess masters in tournaments I can say that the strategies involved in playing chess are much more complicated than medicine. To teach a computer to play chess it is much more than teaching it to solve problems or memorize common moves. I am certain that a nurse with an open access to your textbooks (even if on iphone) could easily get 260+ on step1 or any other medical exam. Because it is all about memorization. But give me access to all the chess strategy books and even let me have a 2nd chess board so I could practice various chess positions and I still wouldn't have a chance against Kasparov. Just like technology would not help me become an artist or a mathematician. By the way if Kasparov did not have a nervous breakdown he would not lose to that computer. His overall game as evaluated by other grandmasters was deemed superior.
Click to expand...

I don't understand how any of you in MS3 and beyond can say that medicine is purely based on memorization. Do you not see the subtle nuances that really drive some diagnoses? Things like reading body language, reading into the way someone speaks, and picking up subtle findings that the patient doesn't even notice play a huge part in medicine. Having a good knowledge base ("memorization" if you will) is definitely important in medicine, but what makes a good clinician is the ability to take that knowledge in combination with how the patient is presenting and synthesize a diagnosis. It's a multifactorial process that requires an understanding- or at least the ability to observe- of human emotions and behavior. Until we have the technological components to observe and understand these things the provider aspect of medicine will not be replaced.

All of you supposed CS people should know that garbage in = garbage out. With something like Watson there still needs to be someone collecting the data for input from the patient- a process which is itself an art.

From the very first day of third year it has been blatantly obvious to me that clinical medicine is so very far from pure memorization. I wonder what you guys are experiencing on your clerkships.
 
yesno said:
You certainly do not need to be a doctor to have good interpersonal communication skills. Businessmen on average have better communication skills...
Click to expand...

Im not talking about interpersonal skills though they are certainly important.

I'm talking about integration of clinical information. Which includes the patient history and interview, physical exam, labs, imaging, etc. integrating information for dx/treatment.

Info that can be looked at in finite values like labs, though they must be considered in context too of hpi/pmh, but things gathered from the interview and exam that does not lend itself to the finite language of computers. Even imaging is a radiologists interpretation and not always definitive.

And the more levels added between a pt history/exam and the doc making a diagnosis leads to an increases likelihood of things being lost. Things are "lost in translation" even when everbody is speaking english. So your world of businessmen with the superb interpersonal skills you think they have would just worsen this. Many doctors often lack these skills, i think more of the old school docs because of more recent trends in training, but its still a different set of skills because of a different interpersonal dynamic then the doctor/patient one. Though some skills are universally helpful in any interpersonal exchange

EMR and computers add safeguards and better recordkeeping but are still limited and will be for at least are lifetime in making docs obsolete if ever. And further tech can enhance clinical scenario for better/quicker treatment outcomes but if anything it will make a docs job harder and require an increased knowledge base. like less simplistic tasks, but a better understanding of molecular genetics to integrate new tests into the clinical scenario. Or old imaging tests replaced by more multifaceted ones that require more of a knowledge base for the radiologist to interpret.

I mean has a physicians job become easier compared to 50 years ago? Absolutely not, the more technology has almost always required an increase in skill and knowledge of the physician to utilize
 
Last edited:
But obviously, making subjective aspects of physicians role more objective and not prone to human error as well as subjective bias is more desirable. I just do not see it being possible to the level physicians become obsolete anytime soon if ever

Do yourself a favor (because its an awesome movie anyway) and watch terminator 2. See the difficulties even the old arnold terminator model has in understanding the human condition and imagine him replacing physicians completely, programmed to just protect like he is john connor and not allowed to kill like john requires. Then realize even that level of learning technology is far from being so prevalent to replace physicians or even exists as far as i know. At best u are still getting a robot physician with the worst interpersonal skills ever and even has a hard to even understanding why humans cry.
 
circulus vitios said:
You guys saying that medicine is impossible for AI are ridiculous. It might be a long way off but it's more than possible once you realize that the human mind is just a biological computer.
Click to expand...

😕

I don't think anyone's arguing that it's impossible. What most of us are saying, however, is that it'll be a long, long, long time before that becomes a possibility. Even then, numerous other professions will be "replaced" by AI before medicine is touched.

circulus vitios said:
Biological circuits and humanistic systems of logic can be transformed to synthetic circuits and man-made software. Our thinking, learning, irrationalities, and intuition are not impossible to recreate.
Click to expand...

We don't even have a solid understanding of how our brain works, what/how our behaviors are influenced, etc. How can you recreate something that you don't fully understand? Considering how complex the human brain is, it'll be a long time before we figure out the intricacies of how it works and it'll be even longer before we can replicate its complexity using AI.
 
Advertisement - Members don't see this ad
I guess I miss-labeled the question by a mile. To steer the discussion a little, from my understanding the role of the doctor is to evaluate the patient (interview him, physical exam, ask questions, etc), come up with a diagnosis (following complex algorithms) and finally prescribe a drug/referral/solution/tell the patient what's wrong with him/her and what his/her options are. these skills and knowledge base come from years and years of memorization/clinical experience.

if we agree that this is the bulk of a physicians skillset - a computer like watson could potentially take over the diagnosis aspect of the duty - where you use natural language to input symptoms, patient history, etc and you get an output of conditions.

here is where I wanted to get your guys' opinions - now if computers can replace this aspect and perhaps do a better job due to its unlimited knowledge bank and lack of bias - i suspect that training time for doctors will decrease and/or salaries as well.

i could be wrong about that last bit - regarding my opinion of what constitutes a physicians "skillset" and would welcome correction.

perhaps this discussion would be best led by someone who had similar concerns when they were premed and now are well into residency/practice.
 
pgoyal said:
I guess I miss-labeled the question by a mile. To steer the discussion a little, from my understanding the role of the doctor is to evaluate the patient (interview him, physical exam, ask questions, etc), come up with a diagnosis (following complex algorithms) and finally prescribe a drug/referral/solution/tell the patient what's wrong with him/her and what his/her options are. these skills and knowledge base come from years and years of memorization/clinical experience.

if we agree that this is the bulk of a physicians skillset - a computer like watson could potentially take over the diagnosis aspect of the duty - where you use natural language to input symptoms, patient history, etc and you get an output of conditions.

here is where I wanted to get your guys' opinions - now if computers can replace this aspect and perhaps do a better job due to its unlimited knowledge bank and lack of bias - i suspect that training time for doctors will decrease and/or salaries as well.

i could be wrong about that last bit - regarding my opinion of what constitutes a physicians "skillset" and would welcome correction.
Click to expand...

Unfortunately, I don't think you understand what a physician's skill-set is. No, it's not simple memorization + following algorithms. Read what dr zaius wrote. It's pretty similar to what I hear from residents and attendings. As I mentioned previously, algorithms will break down at some point -- they are not applicable to every single scenario. And patients often present atypically. Sure, there's lots of evidence for certain things, but there's also a lot of gray area in medicine where clinicians have to make judgments based on their training and experience.

I don't think that a computer can out-perform a full-blown attending any time during our lifetimes. Maybe it'll help with the administrative work, but actually generating accurate diagnoses (consistently) and using clinical judgment? No, that won't happen any time soon.

dr zaius said:
I don't understand how any of you in MS3 and beyond can say that medicine is purely based on memorization. Do you not see the subtle nuances that really drive some diagnoses? Things like reading body language, reading into the way someone speaks, and picking up subtle findings that the patient doesn't even notice play a huge part in medicine. Having a good knowledge base ("memorization" if you will) is definitely important in medicine, but what makes a good clinician is the ability to take that knowledge in combination with how the patient is presenting and synthesize a diagnosis. It's a multifactorial process that requires an understanding- or at least the ability to observe- of human emotions and behavior. Until we have the technological components to observe and understand these things the provider aspect of medicine will not be replaced.

All of you supposed CS people should know that garbage in = garbage out. With something like Watson there still needs to be someone collecting the data for input from the patient- a process which is itself an art.

From the very first day of third year it has been blatantly obvious to me that clinical medicine is so very far from pure memorization. I wonder what you guys are experiencing on your clerkships.
Click to expand...
 
don't know if this was already mentioned on this thread but IBM is sending Watson to the Cleveland Clinic program at Case Western Reserve University School of Medicine..faculty/students will help refine its algorithm for medical decision, it'll be used as a reference tool for PBL, and will take past USMLE questions (past NBME questions)...pretty interesting stuff

http://www.cleveland.com/healthfit/index.ssf/2012/10/ibm_supercomputer_watson_is_he.html

Credit: Article posted from Plain Dealer, October 30, 2012

"CLEVELAND, Ohio -- Watson, the super-smart, Jeopardy!-winning, IBM computer, is headed off to medical school.

In Cleveland.

IBM is sending its precocious computer to the Cleveland Clinic Lerner College of Medicine of Case Western Reserve University.

The announcement was released early this morning by IBM and the Clinic and will be followed up at the Cleveland Clinic Medical Innovation Summit today with a 4 p.m. panel discussion on the role of Watson in the future of healthcare.

Nobody expects the giant thinking machine to graduate in a few years, don a white coat, walk into an operating room and perform surgery.

The goal is to fill Watson's brain with journal articles, conference papers and more medical research than any human could possible cram into his head, then have it come up with a list of possible diagnoses to help doctors make better, faster, more accurate decisions about how to treat their patients.

"It will be a partner, a good sidekick," said Dr. Neil Mehta, director of education technology at Lerner and leader of the project for the school.

"It will be there to help us make sure that we're not missing possibilities, that we're doing a more complete search [of the research] and presenting it in an easy-to-understand manner."

Like it did with questions on Jeopardy!, Watson will tell doctors how confident it is that its diagnoses are correct. And Lerner, Mehta said, will help boost its confidence.

"This is the next big challenge for Watson and IBM," Mehta said. "They have built a second-generation software and they want us to help refine these algorithms to help Watson build its confidence in its answers."

That's not all Watson will do.

It will also provide students with the data it used to get to its answers, which, in turn will help them learn.

Like many of those students, it will conduct research, too.

Lerner medical students will be divided into two groups. A test group that studies with Watson at its side and a control group that studies on its own. And faculty will see which group performs best.

IBM is counting on Watson's group making better decisions and its computer improving medical care.

The technology and consulting company with more than 400,000 employees and $107 billion in revenue in 2011, hopes the Clinic and Watson can work together to create sophisticated software -- a vast reference library that will advance treatment for patients -- that it can sell to healthcare providers around the world.

IBM won't disclose terms, but company officials called this a research-driven collaboration that will advance Watson's ability to help humans make better decisions when it comes to their health.

As smart as Watson is, don't count on it graduating anytime soon.

"Right now we have a three-year plan in front of us," Mehta said. "We probably will be doing some pilot studies early next year."

Those studies will help Watson learn a next-generation technology that IBM calls TeachWatson and improve its Deep Question Answering technology in the field of medicine.

Lerner's 160 students and Watson will pose questions to each other to help the computer become a more analytical thinker, refine its ability to generate hypotheses, rank those hypotheses and come up with answers.

"Every question that is asked is going to help Watson," Mehta said. "He's going to say, 'Oh, this is how humans think.'

"There is so much synergy here because the way students learn right now in the problem-based setting is very similar to the way Watson is being programmed to think."

The New York State-based company chose Lerner over the 170 med schools across the country because it likes its problem-based method of teaching.

"When we want them to learn about the heart and lungs, we give them a case: An elderly gentleman is out shoveling snow and gets chest pain. Students have to work out the problem," explained Dr. James Young, executive dean of Lerner.

That, IBM officials say, is how it wants Watson to think.

"This problem-based learning approach in working with their medical students fits very nicely with the work that we're doing with Watson and the way we want to continue to drive and adapt the technology," said Eric Brown, a manager on the IBM Watson research team.

And Young hopes that will lead to diagnostic suggestions by Watson that are "right on target."

"Obviously Watson's not going to be able to do a cardiac catheterization, but he can clearly say 'This person needs a cardiac catheterization,' " Young said.

"If we can get Watson to give us information in the health-care arena like we've seen with more general sorts of knowledge information, I think it's going to be an extraordinary tool for clinicians and a huge advancement in IT approaches to health care."
 
some of the posts on page 1 were saying watson wasn't designed for medical decision making.. guess IBM is trying to change that..here is the press release by IBM

http://www-03.ibm.com/press/us/en/pressrelease/39243.wss

Cleveland Clinic and IBM Work to Advance Watson’s Use in the Medical Training Field

CLEVELAND, OH- 30 Oct 2012: Cleveland Clinic and IBM (NYSE: IBM) announced today the formation of a collaboration to advance Watson’s use in the medical training field. The IBM team of researchers that created Watson will work with Cleveland Clinic clinicians, faculty and medical students to enhance the capabilities of Watson’s Deep Question Answering technology for the area of medicine.

Watson's ability to analyze the meaning and context of human language and quickly process information to piece together evidence for answers can help healthcare decision makers, such as clinicians, nurses and medical students, unlock important knowledge and facts buried within huge volumes of information.

Watson has been gaining knowledge in the field of medicine, and Cleveland Clinic with IBM recognized the opportunity for Watson to interact with medical students to help explore a wide variety of learning challenges facing the medical industry today.

Rather than attempting to memorize everything in text books and medical journals (now acknowledged as an impossible task), students are learning through doing – taking patient case studies, analyzing them, coming up with hypotheses, and then finding and connecting evidence in reference materials and the latest journals to identify diagnoses and treatment options in the context of medical training.

This process of considering multiple medical factors and discovering and evidencing solution paths in large volumes of data reflects the core capabilities of the Watson technology. Medical students will interact with Watson on challenging cases as part of a problem-based learning curriculum and in hypothetical clinical simulations. A collaborative learning and training tool utilizing the Watson technology will be available to medical students to assist in their education to learn the process of navigating the latest content, suggesting and considering a variety of hypotheses and finding key evidence to support potential answers, diagnoses and possible treatment options.

“Every day, physicians and scientists around the world add more and more information to what I think of as an ever-expanding, global medical library," said C. Martin Harris, M.D., Chief Information Officer of Cleveland Clinic. “Cleveland Clinic's collaboration with IBM is exciting because it offers us the opportunity to teach Watson to 'think' in ways that have the potential to make it a powerful tool in medicine. Technology like this can allow us to leverage that medical library to help train our students and also find new ways to address the public health challenges we face today."

Students will help improve Watson’s language and domain analysis capabilities by judging the evidence it provides and analyzing its answers within the domain of medicine. Through engagement with this education tool and Watson, medical students and Watson will benefit from each other’s strengths and expertise to both learn and improve their collaborative performance. The collaboration will also focus on leveraging Watson to process an electronic medical record (EMR) based on a deep semantic understanding of the content within an EMR.

The shift is clearly away from memorization and towards critical thinking where medical training programs will help student to use powerful discovery and language analysis tools like Watson to help them evaluate medical case scenarios and find evidence to help them carefully rationalize decisions. The physicians will rely on their own experience and expert critical thinking skills to read the evidence and make the final judgments.

“The practice of medicine is changing and so should the way medical students learn. In the real world, medical case scenarios should rely on people’s ability to quickly find and apply the most relevant knowledge. Finding and evaluating multistep paths through the medical literature is required to identify evidence in support of potential diagnoses and treatment options,” said Dr. David Ferrucci, IBM Fellow and Principal Investigator of the Watson project.

Over time, the expectation is that Watson will get “smarter” about medical language and how to assemble good chains of evidence from available content. Students will learn how to focus on critical thinking skills and how to best leverage informational tools like Watson in helping them learn how to diagnose and treat patients.
“New discoveries and medical breakthroughs are growing our collective knowledge of medicine at an unprecedented pace, and tomorrow’s doctors will have to embrace new tools and technology to complement their own knowledge and experience in the field,” said James Stoller, M.D., chair of the Education Institute at Cleveland Clinic. “Technology will never replace the doctor, but it can make us better. Our students and faculty are excited to play a role in getting us there.”

“Being able to work with the faculty and students at an organization like Cleveland Clinic will help us learn how to more efficiently teach and adapt Watson to a new field through interaction with experts. It’s exciting to envision the possibilities for this next phase of Watson,” added Dr. Ferrucci.

IBM and Cleveland Clinic will discuss the role of Watson for the future of healthcare and healthcare education today at the Cleveland Clinic Medical Innovation Summit being held October 29-31, 2012 in Cleveland, OH.

About Cleveland Clinic
Cleveland Clinic is a nonprofit multispecialty academic medical center that integrates clinical and hospital care with research and education. Located in Cleveland, Ohio, it was founded in 1921 by four renowned physicians with a vision of providing outstanding patient care based upon the principles of cooperation, compassion and innovation. Cleveland Clinic has pioneered many medical breakthroughs, including coronary artery bypass surgery and the first face transplant in the United States. U.S.News & World Report consistently names Cleveland Clinic as one of the nation’s best hospitals in its annual “America’s Best Hospitals” survey. About 2,800 full-time salaried physicians and researchers and 11,000 nurses represent 120 medical specialties and subspecialties. Cleveland Clinic Health System includes a main campus near downtown Cleveland, eight community hospitals and 18 Family Health Centers in Northeast Ohio, Cleveland Clinic Florida, the Lou Ruvo Center for Brain Health in Las Vegas, Cleveland Clinic Canada, and opening in 2013, Cleveland Clinic Abu Dhabi. In 2010, there were 4 million visits throughout the Cleveland Clinic health system and 167,000 hospital admissions. Patients came for treatment from every state and from more than 100 countries. Visit us at www.clevelandclinic.org. Follow us at www.twitter.com/ClevelandClinic.

About IBM
For more information, please visit www.research.ibm.com and www.ibmwatson.com.
Registered journalists and bloggers can download video and photos about the Watson and Cleveland Clinic collaboration at The NewsMarket.
 
krafty934 said:
😕

I don't think anyone's arguing that it's impossible. What most of us are saying, however, is that it'll be a long, long, long time before that becomes a possibility. Even then, numerous other professions will be "replaced" by AI before medicine is touched.



We don't even have a solid understanding of how our brain works, what/how our behaviors are influenced, etc. How can you recreate something that you don't fully understand? Considering how complex the human brain is, it'll be a long time before we figure out the intricacies of how it works and it'll be even longer before we can replicate its complexity using AI.
Click to expand...

Exactly

Algorithms is the basis for nursing education and clinical practice not physicians.

only in a few cases, maybe seen more in EM where time to treat is more of a factor would it a physician be following a strict memorized algorithm. Even then an algorithm may need to be augmented for certain patients. Acls, drug OD, poisonings where algorithms were developed by physicians over years of clinical research to get the best outcomes. Watson or any other AI still requires these evidence based outcomes to be programmed into it from work of many physicians and healthcare workers over years and thousands of patients.

You will never have a computer able to push the science of medicine forward as it requires creativity and an inquisitive mind, which even AI or Watson lacks. Although AI computers will have a significant role I do not believe we will ever see a computer able to do further the science of medicine as it will always require.

I mean have at it hoss, the more the merrier but dumping millions and millions into some AI that can at best be a handy tool for pulling information or a $15 hr scribe. While learning from human interaction at the pace of a snail and with a capacity that willl never exceed nor come close to an actual human.

They will never exceed that of a safenet for healthcare workers clinical practice and a data repository that can follow an algorithm based on physician research/innovation.

Standing bet 50 years from now that u AI dreamers will look back on this and realize u put your foot in your mouths.

And no one is arguing how great technology is for medicine, but AI is just like having an artificial rain man. Or like having the most apt pupil ever but one that can never add to the knowledge base it was taught. There are and were plenty of people with higher IQs then einstein, but it took more then intelligence for his achievements. So an AI, watson or whatever can have the highest AIQ ever known but still be as worthless to furthering science as a calculator

Medicine isnt a math problem or a chess game which even the best computers lose to the best chess players. Patients are human beings in a distressed state which only another human can comprehend things like pain, sadness, fear, addiction, manipulation. With the monkey wrench of religious, cultural, and social idiosynchrocies.

Go play d&d nerds
 
Last edited:
topsurgeon2010 said:
some of the posts on page 1 were saying watson wasn't designed for medical decision making.. guess IBM is trying to change that..here is the press release by IBM

http://www-03.ibm.com/press/us/en/pressrelease/39243.wss

Cleveland Clinic and IBM Work to Advance Watson's Use in the Medical Training Field

CLEVELAND, OH- 30 Oct 2012: Cleveland Clinic and IBM (NYSE: IBM) announced today the formation of a collaboration to advance Watson's use in the medical training field. The IBM team of researchers that created Watson will work with Cleveland Clinic clinicians, faculty and medical students to enhance the capabilities of Watson's Deep Question Answering technology for the area of medicine.

Watson's ability to analyze the meaning and context of human language and quickly process information to piece together evidence for answers can help healthcare decision makers, such as clinicians, nurses and medical students, unlock important knowledge and facts buried within huge volumes of information.

Watson has been gaining knowledge in the field of medicine, and Cleveland Clinic with IBM recognized the opportunity for Watson to interact with medical students to help explore a wide variety of learning challenges facing the medical industry today.

Rather than attempting to memorize everything in text books and medical journals (now acknowledged as an impossible task), students are learning through doing – taking patient case studies, analyzing them, coming up with hypotheses, and then finding and connecting evidence in reference materials and the latest journals to identify diagnoses and treatment options in the context of medical training.

This process of considering multiple medical factors and discovering and evidencing solution paths in large volumes of data reflects the core capabilities of the Watson technology. Medical students will interact with Watson on challenging cases as part of a problem-based learning curriculum and in hypothetical clinical simulations. A collaborative learning and training tool utilizing the Watson technology will be available to medical students to assist in their education to learn the process of navigating the latest content, suggesting and considering a variety of hypotheses and finding key evidence to support potential answers, diagnoses and possible treatment options.

"Every day, physicians and scientists around the world add more and more information to what I think of as an ever-expanding, global medical library," said C. Martin Harris, M.D., Chief Information Officer of Cleveland Clinic. "Cleveland Clinic's collaboration with IBM is exciting because it offers us the opportunity to teach Watson to 'think' in ways that have the potential to make it a powerful tool in medicine. Technology like this can allow us to leverage that medical library to help train our students and also find new ways to address the public health challenges we face today."

Students will help improve Watson's language and domain analysis capabilities by judging the evidence it provides and analyzing its answers within the domain of medicine. Through engagement with this education tool and Watson, medical students and Watson will benefit from each other's strengths and expertise to both learn and improve their collaborative performance. The collaboration will also focus on leveraging Watson to process an electronic medical record (EMR) based on a deep semantic understanding of the content within an EMR.

The shift is clearly away from memorization and towards critical thinking where medical training programs will help student to use powerful discovery and language analysis tools like Watson to help them evaluate medical case scenarios and find evidence to help them carefully rationalize decisions. The physicians will rely on their own experience and expert critical thinking skills to read the evidence and make the final judgments.

"The practice of medicine is changing and so should the way medical students learn. In the real world, medical case scenarios should rely on people's ability to quickly find and apply the most relevant knowledge. Finding and evaluating multistep paths through the medical literature is required to identify evidence in support of potential diagnoses and treatment options," said Dr. David Ferrucci, IBM Fellow and Principal Investigator of the Watson project.

Over time, the expectation is that Watson will get "smarter" about medical language and how to assemble good chains of evidence from available content. Students will learn how to focus on critical thinking skills and how to best leverage informational tools like Watson in helping them learn how to diagnose and treat patients.
"New discoveries and medical breakthroughs are growing our collective knowledge of medicine at an unprecedented pace, and tomorrow's doctors will have to embrace new tools and technology to complement their own knowledge and experience in the field," said James Stoller, M.D., chair of the Education Institute at Cleveland Clinic. "Technology will never replace the doctor, but it can make us better. Our students and faculty are excited to play a role in getting us there."

"Being able to work with the faculty and students at an organization like Cleveland Clinic will help us learn how to more efficiently teach and adapt Watson to a new field through interaction with experts. It's exciting to envision the possibilities for this next phase of Watson," added Dr. Ferrucci.

IBM and Cleveland Clinic will discuss the role of Watson for the future of healthcare and healthcare education today at the Cleveland Clinic Medical Innovation Summit being held October 29-31, 2012 in Cleveland, OH.

About Cleveland Clinic
Cleveland Clinic is a nonprofit multispecialty academic medical center that integrates clinical and hospital care with research and education. Located in Cleveland, Ohio, it was founded in 1921 by four renowned physicians with a vision of providing outstanding patient care based upon the principles of cooperation, compassion and innovation. Cleveland Clinic has pioneered many medical breakthroughs, including coronary artery bypass surgery and the first face transplant in the United States. U.S.News & World Report consistently names Cleveland Clinic as one of the nation's best hospitals in its annual "America's Best Hospitals" survey. About 2,800 full-time salaried physicians and researchers and 11,000 nurses represent 120 medical specialties and subspecialties. Cleveland Clinic Health System includes a main campus near downtown Cleveland, eight community hospitals and 18 Family Health Centers in Northeast Ohio, Cleveland Clinic Florida, the Lou Ruvo Center for Brain Health in Las Vegas, Cleveland Clinic Canada, and opening in 2013, Cleveland Clinic Abu Dhabi. In 2010, there were 4 million visits throughout the Cleveland Clinic health system and 167,000 hospital admissions. Patients came for treatment from every state and from more than 100 countries. Visit us at www.clevelandclinic.org. Follow us at www.twitter.com/ClevelandClinic.

About IBM
For more information, please visit www.research.ibm.com and www.ibmwatson.com.
Registered journalists and bloggers can download video and photos about the Watson and Cleveland Clinic collaboration at The NewsMarket.
Click to expand...

Watson basically sounds like a recording device to regurgitate a professors instruction. Sounds like a waste of money for an already overpriced education. The last thing medical education needs is more capital investment
 
jmadden said:
Im not talking about interpersonal skills though they are certainly important.

I'm talking about integration of clinical information. Which includes the patient history and interview, physical exam, labs, imaging, etc. integrating information for dx/treatment.

Info that can be looked at in finite values like labs, though they must be considered in context too of hpi/pmh, but things gathered from the interview and exam that does not lend itself to the finite language of computers. Even imaging is a radiologists interpretation and not always definitive.

And the more levels added between a pt history/exam and the doc making a diagnosis leads to an increases likelihood of things being lost. Things are "lost in translation" even when everbody is speaking english. So your world of businessmen with the superb interpersonal skills you think they have would just worsen this. Many doctors often lack these skills, i think more of the old school docs because of more recent trends in training, but its still a different set of skills because of a different interpersonal dynamic then the doctor/patient one. Though some skills are universally helpful in any interpersonal exchange

EMR and computers add safeguards and better recordkeeping but are still limited and will be for at least are lifetime in making docs obsolete if ever. And further tech can enhance clinical scenario for better/quicker treatment outcomes but if anything it will make a docs job harder and require an increased knowledge base. like less simplistic tasks, but a better understanding of molecular genetics to integrate new tests into the clinical scenario. Or old imaging tests replaced by more multifaceted ones that require more of a knowledge base for the radiologist to interpret.

I mean has a physicians job become easier compared to 50 years ago? Absolutely not, the more technology has almost always required an increase in skill and knowledge of the physician to utilize
Click to expand...

:clap::highfive:
 
It's not an "algorithm". It's pattern recognition in many ways. If you have a patient database with say a billion patients with symptoms, tests, and diagnoses, a computer could look at a patients information and be pretty freaking good at coming up with a diagnosis. It would not forget, get tired, or miss that zebra that has only slightly different systems.

It's not going to happen in the next decade or anything. But it's NOT a flow chart. AIs don't work that way.

Of course, until we get cheap mobile full body scanners that everyone passes through, we will still need someone to do the pt histories and physicals and actually input that data.

The first iteration will probably be just something for physicians to verify their diagnoses and to make sure they haven't missed something. The computer will be wrong frequently. Eventually, it'll be more of a help and people will start being dependant on it. Then likely the insurance companies will require the physicians to submit some extra documentation if they disagree with the computer if they want to be paid. Eventually, that'll occur less and less. I don't know what the profession will look like then. I don't think anyone has anything to worry about right now though - this will take MANY decades. Can't fight progress.
 
someone plugs him back in? :shrug:

Personally I think a Dr. Watson is going to be much more complex than a pattern recognition engine. Diagnostics and even management takes a little more finesse than crunching the numbers and spitting out the best odds. The program will need adjustment region by region, and the ability to recognize when something goes against the odds, and this is beyond the ability of a programmer to understand and therefore well beyond the ability of him to make a computer "understand". Development of something like this will require doctors at every step as well as for use. You could make something that can spit you out the right answer 99% of the time, but this isnt House M.D. The episode doesnt stop once we figure out what it is. If simply identifying an issue based on published patterns was at all effective WebMD would have been more than just a constant headache. Instead it is ineffective to the point of damn near being job security for us.
 
Lokhtar said:
It's not an "algorithm". It's pattern recognition in many ways. If you have a patient database with say a billion patients with symptoms, tests, and diagnoses, a computer could look at a patients information and be pretty freaking good at coming up with a diagnosis. It would not forget, get tired, or miss that zebra that has only slightly different systems.

It's not going to happen in the next decade or anything. But it's NOT a flow chart. AIs don't work that way.

Of course, until we get cheap mobile full body scanners that everyone passes through, we will still need someone to do the pt histories and physicals and actually input that data.

The first iteration will probably be just something for physicians to verify their diagnoses and to make sure they haven't missed something. The computer will be wrong frequently. Eventually, it'll be more of a help and people will start being dependant on it. Then likely the insurance companies will require the physicians to submit some extra documentation if they disagree with the computer if they want to be paid. Eventually, that'll occur less and less. I don't know what the profession will look like then. I don't think anyone has anything to worry about right now though - this will take MANY decades. Can't fight progress.
Click to expand...

The whole problem getting all the variables into the computer because a lot of medicine is subjective data, not objective.

If you think about it, every piece of subjective data can be really broken down into objective data if you have enough time. Everything in the world is 100% cause and effect. However, can we record all those variables in cookbook format? No, it would take far too long.

When we say "a patient looks sick" you could break that down to 500 objective questions which equate to our singular subjective determination but that wouldn't be efficient.
 
Lokhtar said:
You plug it back in?
Click to expand...


You and Specter both missed the point of my question, and I'm sure that's my fault, I was trying to be too clever.

AI proponents say that these machines will replace us. Last I checked, machines require a constant power source to continue functioning. Even in today's world, that's not a given. Look at the hospitals that lost both power and their generators. If they had relied on machines instead of in-house physicians, how many patients would have died? Certainly not all of them, as not everyone needs instant care. Most certainly some would have.

And I already know some of the counter-arguments that I'm sure will come out. I just want to point out that there are limitations to everything, and for any person to claim that something is the be-all-end-all just comes off as fan-boyish to me.


Personally I think a Dr. Watson is going to be much more complex than a pattern recognition engine. Diagnostics and even management takes a little more finesse than crunching the numbers and spitting out the best odds. The program will need adjustment region by region, and the ability to recognize when something goes against the odds, and this is beyond the ability of a programmer to understand and therefore well beyond the ability of him to make a computer "understand". Development of something like this will require doctors at every step as well as for use. You could make something that can spit you out the right answer 99% of the time, but this isnt House M.D. The episode doesnt stop once we figure out what it is. If simply identifying an issue based on published patterns was at all effective WebMD would have been more than just a constant headache. Instead it is ineffective to the point of damn near being job security for us.
Click to expand...

Agree 100%. My pathology professor said something along the lines of "The 'classical' or 'typical' variations of diseases are oftentimes diagnosed less than 50% of the time anymore."
 
NeuroSpeed said:
You and Specter both missed the point of my question, and I'm sure that's my fault, I was trying to be too clever.

AI proponents say that these machines will replace us. Last I checked, machines require a constant power source to continue functioning. Even in today's world, that's not a given. Look at the hospitals that lost both power and their generators. If they had relied on machines instead of in-house physicians, how many patients would have died? Certainly not all of them, as not everyone needs instant care. Most certainly some would have.

And I already know some of the counter-arguments that I'm sure will come out. I just want to point out that there are limitations to everything, and for any person to claim that something is the be-all-end-all just comes off as fan-boyish to me.
Click to expand...

No one is saying is anything is the end all or the be all. We are discussing what's superior. Plus we are talking about diagnosis not treatment. Clearly you'll still need people to put the IVs in and give people medicines - and to make quick decisions when a patient is crashing and situations like that. Robots doing that is quite a bit farther away than Watson doing diagnosis.
 
Advertisement - Members don't see this ad
Plus by the time Watson is that good, it'll run on your phone anyway 😛.
 
Lokhtar said:
No one is saying is anything is the end all or the be all. We are discussing what's superior. Plus we are talking about diagnosis not treatment. Clearly you'll still need people to put the IVs in and give people medicines - and to make quick decisions when a patient is crashing and situations like that. Robots doing that is quite a bit farther away than Watson doing diagnosis.
Click to expand...
Plenty of posters have suggested that robots will eventually be doing surgery and not just medicine, in which case I would hope they could place IVs and administer medications as well.
 
What about people who already distrust medicine thanks to disgraces in medical research? I doubt they'll trust a computer more than a person. There's actually been studies done about how people felt about a robot touching them for procedures; the participants preferred a real person.

I just don't think there will ever be the complete divorcing of humanism in medicine, so I don't believe that physicians will ever be excised from the equation completely.
 
I don't understand why so many people are hung on the idea that computers will replace doctors. What is so special about the medical profession that makes it more vulnerable than other jobs?

If machines can take over medicine, why can't they take over pharmacy, law, accounting, teaching, or even politics? If computers are so smart that they can diagnose and treat, won't they be smart enough to dispense medications, prosecute criminals in courts, and run for president?

Personally, I don't see technology as a threat to medicine, my future profession. In fact, technology should be physicians' best friend. It will increase our efficiency and precision in making diagnoses and delivering treatments. However, we need to keep evolving to meet the need of the advancement in technology, or we will surely be replaced, not by machines, but by more competent humans.
 
I'd say the differential diagnosis is pretty much the easiest thing to do like 80% of the time. Health care providers are not hired purely for their ability to decide what a person has. You show me a robot that can walk into a room, look at someone's boil, touch it, decide if it needs lanced, then does the procedure itself, and I'll be impressed.
 
VenusinFurs said:
I'd say the differential diagnosis is pretty much the easiest thing to do like 80% of the time. Health care providers are not hired purely for their ability to decide what a person has. You show me a robot that can walk into a room, look at someone's boil, touch it, decide if it needs lanced, then does the procedure itself, and I'll be impressed.
Click to expand...

That is trivial from an engineering perspective. Honestly, I could build such a machine with parts in my condo now, maybe another 2-3 thousand dollars worth of stuff and a month off to do some coding. Not sure why you picked an easily doable example. There is a tremendous amount of technical skill in procedural based specialties, but at the end of the day, simple procedures can be taught to just about anyone.
 
mimelim said:
That is trivial from an engineering perspective. Honestly, I could build such a machine with parts in my condo now, maybe another 2-3 thousand dollars worth of stuff and a month off to do some coding. Not sure why you picked an easily doable example. There is a tremendous amount of technical skill in procedural based specialties, but at the end of the day, simple procedures can be taught to just about anyone.
Click to expand...

You could build a walking, diagnosing, procedure-giving autonomous robot with parts around your condo, a couple thousand in cash, and a month of time? Tony Stark, is that you? I think you went into the wrong field if all that is true.
 
Renaissance Man said:
You could build a walking, diagnosing, procedure-giving autonomous robot with parts around your condo, a couple thousand in cash, and a month of time? Tony Stark, is that you? I think you went into the wrong field if all that is true.
Click to expand...

Nobody said anything about diagnosing. The poster that I quoted said that that was the easy part and stated that the physical procedure of identifying if a boil needs to be lanced and then lancing it was the hard part. The most expensive part would be the ultrasound transducer, but other than that, it isn't exactly rocket science to to design something to drain a superficial abscess accurately.
 
mimelim said:
Nobody said anything about diagnosing. The poster that I quoted said that that was the easy part and stated that the physical procedure of identifying if a boil needs to be lanced and then lancing it was the hard part. The most expensive part would be the ultrasound transducer, but other than that, it isn't exactly rocket science to to design something to drain a superficial abscess accurately.
Click to expand...

I think the point he was making was generating a diagnosis machine alone will not suffice and robotics is further away from making a machine that can do a meaningful examination than it is from making something that can produce diagnoses based on inputs. Doing both for itself (performing the exam, interpreting the exam itself, and making a judgement on what to do, then carrying it out) is even further.

The only practical way for your hypothetical condo machine to be implemented will still require someone to make the call on whether or not to use your mechanical boil lancer which isnt really the point of this thread.

Sent from my DROID RAZR using SDN Mobile
 
Last edited:
mimelim said:
That is trivial from an engineering perspective. Honestly, I could build such a machine with parts in my condo now, maybe another 2-3 thousand dollars worth of stuff and a month off to do some coding. Not sure why you picked an easily doable example. There is a tremendous amount of technical skill in procedural based specialties, but at the end of the day, simple procedures can be taught to just about anyone.
Click to expand...

We should get a kickstarter campaign going for you. Anyone who donates more than $100 gets a free boil lancing when you build your robot.
 
mimelim said:
Nobody said anything about diagnosing. The poster that I quoted said that that was the easy part and stated that the physical procedure of identifying if a boil needs to be lanced and then lancing it was the hard part. The most expensive part would be the ultrasound transducer, but other than that, it isn't exactly rocket science to to design something to drain a superficial abscess accurately.
Click to expand...

What I'm saying is that the robot being able to decide whether to lance or not to lance would be the tricky part. A human looking at a boil would poke at it and kind of decide if it was big and pus-filled enough to try to drain. I don't know that you could easily design a robot to do that. Anyone could figure out that a large fluctuant skin nodule is probably an abscess. Getting the robot to get to determine that something is a large fluctuant skin nodule is the tricky part.

Also I don't believe you, but that's a separate thing entirely.
 
VenusinFurs said:
What I'm saying is that the robot being able to decide whether to lance or not to lance would be the tricky part. A human looking at a boil would poke at it and kind of decide if it was big and pus-filled enough to try to drain. I don't know that you could easily design a robot to do that. Anyone could figure out that a large fluctuant skin nodule is probably an abscess. Getting the robot to get to determine that something is a large fluctuant skin nodule is the tricky part.

Also I don't believe you, but that's a separate thing entirely.
Click to expand...

I don't think it would be difficult using a pressure sensor and ultrasound.
 
VenusinFurs said:
What I'm saying is that the robot being able to decide whether to lance or not to lance would be the tricky part. A human looking at a boil would poke at it and kind of decide if it was big and pus-filled enough to try to drain. I don't know that you could easily design a robot to do that. Anyone could figure out that a large fluctuant skin nodule is probably an abscess. Getting the robot to get to determine that something is a large fluctuant skin nodule is the tricky part.

Also I don't believe you, but that's a separate thing entirely.
Click to expand...

No it's not. You would build a classifier, i.e. a simple neural network where the inputs are the same characteristics that humans use to identify a "large fluctuant skin nodule". You would start with an image of the lesion to look for color, shape, size, texture etc. A human would also "poke at it", which could be accomplished with ultrasound/pressure transducer as mimelim already mentioned. Computer science undergrads with pattern recognition experience could code this in an afternoon.

TheProwler said:
You could write the software to recognize the fluid cavity? using what programming language?
Click to expand...

Pick one. It doesn't matter what language. Personally, I would use matlab.
 
Advertisement - Members don't see this ad
TheProwler said:
You could write the software to recognize the fluid cavity? using what programming language?
Click to expand...

While in undergrad I wrote java based programs to outline and do 3d reconstructions of tumors on ultrasounds. Initially we just manually used the probes to grab out images, but I pushed them to let me adapt a C&C machine to more systematically scan a portion of the body.

In high school I designed a robot as part of a competition to automatically find a grape suspended in jello and then skewer it with a rod without puncturing the back edge of the grape. Given that we did that using legos and crappy proprietary software to do that, I am sure this can easily be done from the ground up.

Again, I am in the, "MDs won't be replaced by computers/robots ever" crowd. I was simply pointing out that this is a relatively trivial indication. It would not take very many resources or even time to do this. Given that I am a relatively poor programmer and I can see how to do this, I am sure that there are hundreds of people out there that could do this if someone sat down and explained exactly what they needed.
 
mimelim said:
While in undergrad I wrote java based programs to outline and do 3d reconstructions of tumors on ultrasounds. Initially we just manually used the probes to grab out images, but I pushed them to let me adapt a C&C machine to more systematically scan a portion of the body.

In high school I designed a robot as part of a competition to automatically find a grape suspended in jello and then skewer it with a rod without puncturing the back edge of the grape. Given that we did that using legos and crappy proprietary software to do that, I am sure this can easily be done from the ground up.

Again, I am in the, "MDs won't be replaced by computers/robots ever" crowd. I was simply pointing out that this is a relatively trivial indication. It would not take very many resources or even time to do this. Given that I am a relatively poor programmer and I can see how to do this, I am sure that there are hundreds of people out there that could do this if someone sat down and explained exactly what they needed.
Click to expand...

Did this jello submerged grape skewering machine have a purpose, or was it just to show the precise nature with which you could code a robot to use pressure?
 
http://www.kevinmd.com/blog/2011/09/learning-hard-art-medicine.html

do you guys think a watson like technology could have helped here? http://www.kevinmd.com/blog/2011/02/watson-board-certification-exams-obsolete-medicine.html

just wondering why physicians don't input data like these onto a "software" to track when and how they went wrong with the diagnosis. this way they could learn from their mistakes and also enable other physicians to be on the lookout for this. (disregard if this already happens)
 
i highly recommend the following link & video. might add to the discussion and educate some more about the true capabilities of watson.

http://www.youtube.com/watch?v=3G2H3DZ8rNc

http://www.kevinmd.com/blog/2011/08/patients-trust-sociable-humanoid-robots.html
http://www.kevinmd.com/blog/2011/02/watson-board-certification-exams-obsolete-medicine.html


i'd like to steer this discussion toward a "pure speculation out of interest on the future of medicine and role of doctors/technology affecting doctor pay" type...and avoid attacking one another if possible. i think all opinions are valid and some reasoning/"evidence&examples" behind them would be appreciated. friendly discussion 🙂
 
mimelim said:
While in undergrad I wrote java based programs to outline and do 3d reconstructions of tumors on ultrasounds. Initially we just manually used the probes to grab out images, but I pushed them to let me adapt a C&C machine to more systematically scan a portion of the body.

In high school I designed a robot as part of a competition to automatically find a grape suspended in jello and then skewer it with a rod without puncturing the back edge of the grape. Given that we did that using legos and crappy proprietary software to do that, I am sure this can easily be done from the ground up.

Again, I am in the, "MDs won't be replaced by computers/robots ever" crowd. I was simply pointing out that this is a relatively trivial indication. It would not take very many resources or even time to do this. Given that I am a relatively poor programmer and I can see how to do this, I am sure that there are hundreds of people out there that could do this if someone sat down and explained exactly what they needed.
Click to expand...

Yeah, but do you know how to take a screenshot using a Mac like I do?

That is pretty awesome though, still sounds like you should start developing medical equipment for some side money.
 
You must log in or register to reply here.