×

Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

Before you choose to head back to the Classic look of the site, we'd appreciate it if you share your thoughts on the Beta; your feedback is what drives our ongoing development.

Beta is different and we value you taking the time to try it out. Please take a look at the changes we've made in Beta and  learn more about it. Thanks for reading, and for making the site better!

Watson Goes To Medical School

samzenpus posted about a year and a half ago | from the computer-learning dept.

IBM 100

First time accepted submitter Kwyj1b0 writes "I.B.M's Watson is headed to the Cleavland Clinic Lerner College of Medicine of Case Western Reserve University for training. Clinicians and students will answer and correct Watson's questions, in an attempt to crowdsource its education. From the article: '“Hopefully, we can contribute to the training of this technology,” said Dr. James K. Stoller, chairman of the Education Institute at Cleveland Clinic. The goal, he added, was for Watson to become a “very smart assistant.” Part of Watson’s training will be to feed it test questions from the United States Medical Licensing Exam, which every human student must pass to become a practicing physician. The benefit for Watson should be to have a difficult but measurable set of questions on which to measure the progress of its machine-learning technology.'"

cancel ×
This is a preview of your comment

No Comment Title Entered

Anonymous Coward 1 minute ago

No Comment Entered

100 comments

What other urgent jobs do editors have? (0)

Anonymous Coward | about a year and a half ago | (#41836607)

that they can't even bother Cleveland? Do you guys tweet your submissions from the bathroom?

Re:What other urgent jobs do editors have? (2, Funny)

Anonymous Coward | about a year and a half ago | (#41836631)

oops, looks like I couldn't bother verb there. I should get a /. editor job too.

Re:What other urgent jobs do editors have? (0, Troll)

Anonymous Coward | about a year and a half ago | (#41836693)

Also, I'm a gay fagget that still can't figure out what verb I was trying put in the sentence.

Re:What other urgent jobs do editors have? (0, Troll)

Anonymous Coward | about a year and a half ago | (#41836711)

Gay fagget means you're straight or a bunch of gay sticks? You didn't spell faggot right, faggot.

Watson - not for vets! (0)

Anonymous Coward | about a year and a half ago | (#41836681)

So, I have been following Watson and it's inevitable turn to the medical profession. They tried simple A.I.-based programs in the 1980's I believe, but it never really took off.

I work for one of the largest corporations in the veterinary industry, so I get the chance to talk to veterinarians about their computer systems often. We sell a product that combines different aspects of the business on one platform; billing, scheduling, medical history, inventory management, reporting, and more. If something like Watson is available to the veterinary industry, we would be the first to sell it. While most like the ability for the computer system to generate treatment plans for patients (patient has history of X and Y, so increased chance of Z developing, but plans A or B can mitigate the risk of developing issue Z), the majority of them (and this is all anecdotal evidence, mind you) would *never* want a program to tell them how to do their jobs. I understand that that isn't *exactly* what Watson does or will do, but its pretty close.

Besides, what happens when Watson realizes that complete extermination of homo sapiens is the only solution to solve our health problems?

captcha - "compute"

Re:Watson - not for vets! (4, Insightful)

lattyware (934246) | about a year and a half ago | (#41836837)

Watson actually does some really cool stuff with respect to not just being 'do this' - it tells you what it 'thinks', but also tells you why it thinks those things, and how sure it is of that. So it will say 'I think it's likely the patient has Y because of family history, environmental factors, this symptom and these studies', etc... It's more giving the doctor all the (relevant) information possible than telling the doctor what to do.

Re:Watson - not for vets! (1)

WillAffleckUW (858324) | about a year and a half ago | (#41837105)

In the medical profession we call those things Lists. A number of studies, based on original research at the UW Medical Center and VA hospitals, have found that check lists do wonders.

But ... we already do that. We even have iPads in VA Hospitals now. Watson is just doing checklists.

Re:Watson - not for vets! (0)

Anonymous Coward | about a year and a half ago | (#41837711)

Sure Watson is just doing checklists like a motorized boat is just rowing.

Re:Watson - not for vets! (0)

Anonymous Coward | about a year and a half ago | (#41840509)

Sure Watson is just doing checklists like a motorized boat is just rowing.

But does it doe checklist like a submarine is swimming?

Re:Watson - not for vets! (1)

MysteriousPreacher (702266) | about a year and a half ago | (#41839623)

Sure, and I could get a team to spend a week performing analysis to determine appropriate staffing levels - dealing with tens of thousands of data points, with data at all sites varying on an hourly basis. Alternatively I could have a learning system that keeps far better track of these things than humans would, and presents its working for review before implementation.

Re:Watson - not for vets! (1)

TapeCutter (624760) | about a year and a half ago | (#41840365)

"Watson is just doing checklists" - You miss the point, in a year or two, it will probably be more capable of writing and correcting those lists than any single human being.

Re:Watson - not for vets! (0)

Anonymous Coward | about a year and a half ago | (#41837553)

Who cares? Programmed by people, Watson will only work as well as the people who programmed him.

Get back to me when computer have their own intelligence without being programmed.

Re:Watson - not for vets! (4, Interesting)

Unnngh! (731758) | about a year and a half ago | (#41837821)

You are wrong, broadly speaking. This is the whole point of machine learning: given a very complicated task that it would take a human a tremendous amount of effort to program correctly, you can instead get the machine to figure out how to perform the task itself, rather than explicitly programming it to do one thing. Some types of learning are supervised, particularly classifiers: I tell the computer which items belong to which class, and given a new, previously unseen item, the computer attempts to determine its class based on the training. Others are unsupervised: set the robot free in the environment with some goal function and let it learn through trial-and-error how to optimize its behavior toward the goal. Watson is a combination of first-order logic (prolog and a huge kb) and a variety of such learning algorithms. Some of this is stuff that was considered an industry failure in the 80s but, paired with modern machine learning techniques, is quite powerful. Indeed we may be seeing the first instances of computers that have some form of this "intelligence" of which you speak, though I think we are still a long way from "strong AI".

Re:Watson - not for vets! (1)

WayfinderSteve (2659663) | about a year and a half ago | (#41846955)

Watson better learn to read crabby handwriting and minds. The fundamental issue in applying AI to medical data is the low information dataset. We are working on a couple of smaller AI projects in my hospital and finding that even when electronic, most entries in the record are narrative not discrete.

Applying ranges and logic to data pulled out of narrative records is tricky and leads to unusual responses. Even when discrete, the data set may be difficult to use. My favorite example seen nationally is blood transfusion data. It can be in units of 250ML, 500ML or just ML. Users don't look at the unit label for a field. If we ask for the most precise measurement in ML, we get amounts of 1 or 2. We know nobody got a thimble of blood.

AI interventions have to be quick. It's no good telling the provider later he made a mistake. The goal is to steer them in the right direction before they act without presenting them with so much info they throw in the towel. Not easy and in infancy.

Re:Watson - not for vets! (0)

Anonymous Coward | about a year and a half ago | (#41849411)

Great points - cleaning the data sets is usually one of the hardest problems when practically implementing AI for any moderately interesting data. I can see how the medical field would be a sort of worst case scenario for this. The little medical data I've seen, from controlled studies, was a huge mess, with no way to reconcile things like ML as you mentioned, and often no way to even tell what type of thing the fields were meant for to begin with. It is still a really hard problem.

Re:Watson - not for vets! (1)

wmac1 (2478314) | about a year and a half ago | (#41839091)

Older systems used expert systems (hierarchical knowledge systems). The new models use machine learning (neural network classifiers, ...).

The machine learning creates a generalized model based on small amount of data. Expert systems search the existing data and the data should be very accurate.

Cleveland, not Cleavland (0)

Anonymous Coward | about a year and a half ago | (#41836683)

Who edits this stuff?

Re:Cleveland, not Cleavland (0)

Anonymous Coward | about a year and a half ago | (#41841715)

Self-driving cars.

They are (rightly) really proud of this at IBM. (4, Interesting)

lattyware (934246) | about a year and a half ago | (#41836703)

I recently worked at IBM and this is one of the things they really love to showcase - I think primarily because it's really cool, but also has really useful end results - exactly the kind of thing you want to be working on.

but standardized tests are not knowledge. (0)

Anonymous Coward | about a year and a half ago | (#41837959)

As was pointed out up-thread, we already have checklists. A better checklist is a big deal, but where is the secret sauce?

A student taking a standardized test may have knowledge and the test may be a limited view of the knowledge the student may possess, which helps quantify the student's level of competence. But, teaching to the test just means the students can provide answers to a limited set of questions, there is no depth. And with computers, more so.

While it may be naive to expect a computer to 'understand' the subject matter, I would hope that the people using the tool are not so confused that they mistake the ability to respond correctly to a set of test questions with actual command of the field.

OK, let's read your aura now. :-()

IBM is all gimmick and no substance (-1)

Anonymous Coward | about a year and a half ago | (#41838469)

It's purely a marketing gimmick, it will never be good enough to even run WebMD. When Watson won Jeopardy, Watson had the text of the questions fed to it at the same time the question was being read out to the contestants, whereas the people playing had to do voice recognition like Siri does. WATSON COULDN"T EVEN WIN SIRI AT JEOPARDY, because in the real world it would have to do voice recognition too and Siri can do that and Watson cannot.

We see the same thing with their processors, they launch the z196 clocked at it's fastest and pretend that fastest clocking means fastest processor! But there processors are abysmally slow! Several orders of magnitude below the fastest production processors, and they won't even let you benchmark them to prevent comparison. However occasionally someone does anyway, and the results are not pretty:

Z114 clocks in about 1200 MIPS for the top of the range, the Intel chips are more than 100x that. Their $75k system only delivers 29 MIPS....

http://www.tech-news.com/publib/pl2818.html

But that's OK for them as long as they can keep doing gimmicks. They'll parade Watson around for a while, patent everything sites like WebMD do right now, then sue future companies based on those patents. Gimmicks.

Re:IBM is all gimmick and no substance (0)

Anonymous Coward | about a year and a half ago | (#41840773)

Obviously you are not a programmer, nor do you have any knowledge of machine learning... Siri is a command interpreter. "What is the weather like outside?" "What Time is it?" Those are questions that have definitive answers and take no amount of learning to solve. For those questions it only has to figure out a few things. 1) where you are, 2) how to load a weather website for that location 3) how to read the system clock... and those things are all functions in the program for performing a specific task. You will never find anything even remotely close to that written into Watson since all of it's information including what answers it can figure out are stored in it's learned database. WebMD is a lookup table for symptoms. I can type in symptoms, and it will give me suggestions and let me read through the actual material to determine if that's what I have. Watson looks at more than just symptoms. Part of it's advantage is that is has more information, and part of it is the techniques it has for looking up answers in the information it has. Not just a keyword search as you seem to think. Maybe you didn't actually watch the Jeopardy episode.
Also, it is obvious you have no real knowledge of the intent of Z processors or how they work either... A 1-1 comparison with an intel chip isn't really possible since the processors have quite different architectures, are designed to solve very different problems, and have very different duty cycles. A Z chip is designed to run with extremely high reliability, and contain many more circuits to make sure the answer given is the correct answer. So it will never make a mistake, and will even be able to tolerate some degree of hardware failure before shutting down and alerting the system, all while running at a much higher duty cycle than any intel processor could handle. The processor in the new Mars Rover is quite slow compared to the processor in your desktop or laptop, but cost in the neighborhood of $20k each... Why is that?? Well... a few reasons. !) They didn't make a lot of them, 2) There is a lot more radiation to contend with so the processor actually runs each instruction through multiple circuits simultaneously and then votes on the answer... So your super fast intel chip would be dead in a second up there. Glad you aren't in charge of that project...

Gimmick, caused by Patent Appeal Court (0)

Anonymous Coward | about a year and a half ago | (#41839337)

It's a gimmick.

There are symptom diagnosis software already, there is natural text language parsers already. There are companies that have joined the two together and found they don't work well enough to avoid being sued for medical malpractice.

However the patent appeals court made it possible to patent X+Y, even if X exists and Y exists and X+Y is obvious. It overruled the Supreme court, that said the invention must be more than the sum of its parts. So now you can simply bolt two things together and if it doesn't exist already you can patent it.

Oh course nobody has ever made a commercial Language parser + diagnosis because it would be too risky, you'd get sued too much. But that doesn't matter to IBM, because IBM has no plans to make one that works. It only plans to patent in that area, then wait for someone else to solve the actual problems, and take their money.

I don't think IBM staff should be proud of these marketing patent stunts. They should be proud when IBM makes something in a competitive market and wins by virtue of being the better product. Something sadly almost non existent for IBM these days.

Re:They are (rightly) really proud of this at IBM. (0)

Anonymous Coward | about a year and a half ago | (#41843705)

How long before I can get an Ask Watson app for my iPhone? That bitch Siri is dumb as rocks.

Repurpsose for Criminology (5, Funny)

p0p0 (1841106) | about a year and a half ago | (#41836733)

Design another Watson designed as a database for crimes which could analyse crime scenes, point out potential minute details and give data on similar crimes. Then call it... Holmes.
I'd like that.

Re:Repurpsose for Criminology (1)

lattyware (934246) | about a year and a half ago | (#41836909)

Just as a note, you wouldn't need another Watson, just run it again, and train it on different data. It could (in theory) apply it's 'learning' skills to anything.

Re:Repurpsose for Criminology (-1)

Anonymous Coward | about a year and a half ago | (#41836943)

Don't tell us nothing, shitball. Just shut your fucking mouth. No one cares.

Re:Repurpsose for Criminology (1)

Anonymous Coward | about a year and a half ago | (#41837005)

Even better, set up multiple Watsons, have them work over the same sets of data to learn from, and then ask them all the same set of questions, and see how each system answers those questions...

Re:Repurpsose for Criminology (0)

Anonymous Coward | about a year and a half ago | (#41839813)

The same.

More interestingly is to train them on different sets of data, then get them to vote on answers. This works exceedingly well for random forests.

Re:Repurpsose for Criminology (2)

p0p0 (1841106) | about a year and a half ago | (#41837133)

Then you couldn't put them in the same room and crudely re-enact scenes from Sherlock Holmes.
And a little USB drive shaped like a pipe.
Yeah. That would be awesome.

Invert the logic... (1)

Anonymous Coward | about a year and a half ago | (#41837331)

...and call it Professor Moriarty

Holmes here (1)

Anonymous Coward | about a year and a half ago | (#41838727)

Nurse: Doctor patient Parkison, has numbness in the groin area.
Web MD : Did the numbness come on suddenly?
Nurse: No.
Web MD: Does it affect both sides?
Nurse: Yes.
Web MD: Thoracic Spinal Stenosis

Watson: Parkinson's disease. (Patent Pending)

Really they've put the text through a crude parser and stuck it into a database similar to Web MD and now they'll patent the hell out of this. But they're using stock language parsing algorithms and stock medical look databases, and the only purpose for this is to generate fake cover for a lot of patents in the medical field.

You won't see any product from IBM in this field other than the patent.

Re:Repurpsose for Criminology (1)

houghi (78078) | about a year and a half ago | (#41839509)

After a while, make it more advanced and let it predict crimes [wikipedia.org] , so we can prevent them. Have three and call them Mike, Donna and Jerry (Please not Agatha, Dashiell or Arthur.)

Sci-fi stuff (1)

BumpyCarrot (775949) | about a year and a half ago | (#41836765)

Reminds me of a short story from the Stories of Ibis collection. A salesman impressed a group of colleagues in a nursing home that a new robot had passed the nursing exam.

What is Toronto?????? (1, Troll)

Joe_Dragon (2206452) | about a year and a half ago | (#41836789)

just wait for it to make a error like that in a Medical setting.

Re:What is Toronto?????? (0)

lattyware (934246) | about a year and a half ago | (#41836829)

And that's why you still have actual doctors - this is simply a way to get doctors information quickly and filter the relevant stuff down, not make decisions for them.

Re:What is Toronto?????? (0, Redundant)

geekoid (135745) | about a year and a half ago | (#41836839)

Someday it will replace doctors.

Re:What is Toronto?????? (1)

TheLink (130905) | about a year and a half ago | (#41836887)

They'll find other jobs, just like the buggy whip makers.

That's what the free market capitalists like to say anyway.

but under the mitt romney no health care plan (1)

Joe_Dragon (2206452) | about a year and a half ago | (#41836953)

Errors like that can get you on the pre-existing condition black list. Where after that your only doctor may be Watson ER

Re:but under the mitt romney no health care plan (0)

Anonymous Coward | about a year and a half ago | (#41837015)

Or we can go with Obamy's plan and you'll never see a doctor because of the backlog.

Cause that is the result of private insurance? (0)

Anonymous Coward | about a year and a half ago | (#41837129)

That's all the Affordable Care Act is essentially: premium support. Turn off the talking heads, read.

Such a beast is sure to be a Microsoft product. (0)

Anonymous Coward | about a year and a half ago | (#41837031)

A poor clone of Watson, not the original. This is so obvious, it's elementary.

Re:What is Toronto?????? (1)

Yvanhoe (564877) | about a year and a half ago | (#41837013)

Human doctors fail as well. We will judge Watson on a criterion of 0% of errors, but given the rate of malpractice in US, having a device able to cut by 2 would be a huge progress.

Re:What is Toronto?????? (0)

Anonymous Coward | about a year and a half ago | (#41838723)

Doctors make errors every day, too. In fact, most of the time they use their intuition to put the diagnostic. Also, doctors can't keep up with the latest research and sometimes are just not competent enough, or can't take all the details in the head at once. If the problem is more complex than the brain, then there is no chance to solve it.

Another annoying buzzword getting beaten to death (0)

Anonymous Coward | about a year and a half ago | (#41836797)

Crowdsourcing.
 
So now everything done by a group of people is suddenly crowdsourcing? What's up with the trend to put the word "source" in every phrase?
 
It's getting annoying.

“very smart assistant” (0)

Anonymous Coward | about a year and a half ago | (#41836821)

Ha! Suckers! (cue evil laugh). You will be Watson's assistant, once it has sucked your brain dry. We will call you Igor.

:-)

Robots as doctors (0)

Anonymous Coward | about a year and a half ago | (#41836843)

Let's hope this goes better than internet websites such as WebMD :

http://www.smbc-comics.com/index.php?db=comics&id=2291#comic

Re:Robots as doctors (1)

lattyware (934246) | about a year and a half ago | (#41836881)

I know it's just a joke, but just to clarify: the big difference here is the scope of what Watson does. Websites generally take one symptom and list off all possible causes (so naturally people focus on the worst ones). Watson takes (in natural language) patient history, symptoms, environmental factors, current medication, etc... - everything a doctor has access to (and a doctor would have to perform the exam to make this worthwhile), and then cross references with a huge amount of knowledge from papers, classes like this, 'experience', etc... this is then used to give an idea of what the patient has, how to treat, etc... It's really cool.

Could be great for medical diagnosis (3, Insightful)

Sycraft-fu (314770) | about a year and a half ago | (#41836853)

Particularly since the way it work is by probabilities. So a physician goes and inputs all the symptoms a patient reports, perhaps along with a confidence of how likely it is to be real. Watson could then spit out the likely causes, and the probability of each, as well as how to narrow it down. Then with additional tests, they can exclude things and get a re-factored list.

It won't remove the need for a medical professional with good judgement, but it could be a boon for searching through things and presenting possibilities. What's more each new case can be logged, improving its database.

So when someone presents with a rare disease, it would be much easier for a physician to diagnose it, even if they've never heard of it.

If implemented right, it could cut down on misdiagnosis a ton.

Re:Could be great for medical diagnosis (2)

lattyware (934246) | about a year and a half ago | (#41836939)

Watson is actually already cooler than that, you just give it the patient records, and results from the exam you just did, for example (all in natural language, no special inputting with confidence or anything like that), and Watson will take everything into account (from symptoms, environmental stuff, existing medication, patient history, family history, etc...) and give causes, treatments, etc... - it's really awesome.

Re:Could be great for medical diagnosis (1)

Empiric (675968) | about a year and a half ago | (#41837735)

*citation needed

Whenever I hear anthropomorphizing phrases like "will take everything into account", my overstated-AI BS alarm goes off.

Can you elaborate on by what means or personal experience you assert that Watson does more than statistical analysis of the language it is "given", as essentially-arbitrary symbols, and, say, give some means (or even a data structure) by which it "knows" even what a "patient" is, such that it could draw inferences about such an entity in the absence of an existing chain of words providing the answer via textual comparison of the terms and/or synonymous symbols?

Re:Could be great for medical diagnosis (0)

Anonymous Coward | about a year and a half ago | (#41838795)

Do people do more than that?

Re:Could be great for medical diagnosis (1)

Empiric (675968) | about a year and a half ago | (#41839161)

Yes, for one, people can do "in a similar manner as" inferences that tend to be rather vicious things to represent flexibly enough in data or write a general algorithm for.

You may have seen such things in IQ tests for people, the "X is to Y as A is to B" questions...

"Tree is to forest as book is to _______".

It is not difficult for a human to pick an answer "library" from a multiple-choice list including, say, "cover", "magazine", and "pages". Having a computer score as well on a series of these, though, is to my mind, something as difficult to do as reliably passing the Turing Test, and much of the capability of human intelligence (and a hypothetical true AI system) is provided by the ability to do thought processes like this.

I know of no implementation that has a reasonable hope of meeting this objective that exists, and maybe I was a bit harsh with the GP in expressing that. Being overly generous in accepting claims of a product being AI is easy to do given the complexity of the field, and such claims have a lot financial incentive to be made falsely by companies. I remember a nationwide TV ad campaign 10-15 years ago by, IIRC, Compuware, touting their "software that thinks". The only problem is that it did, and does, no such thing, and the claim serves only to cheapen legitimate accomplishments of the field for deceptive financial gain. Seeing a procession of similar misrepresentations over time (e.g. Siri), it's become something of a "pet annoyance" of mine.

Re:Could be great for medical diagnosis (0)

Anonymous Coward | about a year and a half ago | (#41842613)

Considering that Watson won a game of Jeopardy, it might be capable of solving SAT style analogies as well.

Re:Could be great for medical diagnosis (1)

Empiric (675968) | about a year and a half ago | (#41847953)

Then present some evidence of that. Solving analogies is much, much harder than guessing at a probable appropriate word to pick (prefixed with "Who/what is a...") after running the question string into Google and doing a statistical analysis of the words in the Top 1000 search results.

Re:Could be great for medical diagnosis (0)

Anonymous Coward | about a year and a half ago | (#41839175)

Is Winston a man or a woman? If you're in Brixton what ethnicity is he and how many fingers+toes will he have on average at birth? (Caribbean descent has a common mutation, an extra digit which is removed surgically).

Let me put it more bluntly. IBM will never make either a natural language parser, or a medical lookup system from this. Their previous attempts at complex technology have not gone well, Via Voice being a typical example. It was crap in the market, however they did create a lot of patents from it and those patents let them milk other companies who *did* make good voice systems. To hold a patent they have to look like a technology company without ever having to sell a product in the market in competition to others.

Watson is simply marketing, it's always applied to limited context problems that are trivial to solve. Jeopardy was a classic example, on the face of it, it's a general knowledge quiz, but in actuality it uses only a few formats for phrasing a question (or rather answer), and all the questions can neatly fit into a small simple box without any vague subjectivity. A perfect limited subset that existing systems can easily do, that will appear to be clever.

No other expert system will fight it out on Jeopardy, just as this will never be allowed to be compared to the mature diagnosis software already out there. Once its served its purpose of marketing IBM and letting them claim a lot of patents, they'll move it onto some other area.

Re:Could be great for medical diagnosis (0)

Anonymous Coward | about a year and a half ago | (#41837171)

Like a Bayesian Network... which have been used for exactly that for a long time.. Watson is cool because of the Natural Language Processing aspects and its ability to handle large unstructured data.

Re:Could be great for medical diagnosis (0)

Anonymous Coward | about a year and a half ago | (#41837803)

we already have this in pediatrics (sans the physician indicating a degree of confidence in a given finding). It is called simulconsult and helps develop a differential diagnosis for complex presentations so the physician can begin a diagnostic work up with the goal of narrowing things to the point where, if needed, consultation with the right human expert is possible.

Re:Could be great for medical diagnosis (0)

Anonymous Coward | about a year and a half ago | (#41838181)

Idiocracy, plug the Watson tube into the patient and watch the display.

Re:Could be great for medical diagnosis (1)

EnsilZah (575600) | about a year and a half ago | (#41838399)

More than diagnosing the disease based on the symptoms, if you had a networked system available in every hospital or to every doctor, I'd like to see what kind of information you'd get out of following up with post-treatment symptoms, autopsies when things went wrong, detecting patterns that wouldn't be evident on a per-doctor or per-hospital scale.
Of course I'd imagine you'd need to hash the input case numbers or something to keep the privacy of the individuals while being able to add information later.
Also I wonder how a system like this would scale to serve multiple hospitals.
Would it have to store information gathered during the rush hours and then search for new patterns later, sort of like what the suggested purpose of dreaming is?

Re:Could be great for medical diagnosis (1)

houghi (78078) | about a year and a half ago | (#41839599)

Well, to be diagnosed with a rare disease, you have to be diagnosed first to have that rare disease and not a common one.
Doctors will not start looking for a rare disease if they think you have a common one. They might start looking once they realize you don't have a rare disease.

Even having a common one confused with something else happens more then once.

I have gout and it took a long time before they realized it was gout. When it flares up, I have it in my ankle and it was thought the first few times that it was cause by twisting it. This because that happened very often when I was a kid.

Yes, I had seen several doctors and they all came up with the wrong conclusion. Even now when I have an attack and go to a different doctor, I must say specifically that I did not twist it and that I am already diagnosed with gout.

And no, doctors will not start using this system each and every time they get a patient. Yes, there will be cases where they have no idea what you have and then this might help.

Oh, concerning the milk if it was bad or good, my doctor could not say it from the top of his head to be 100% sure and he used google. I know that Wikipedia says "The consumption of coffee, vitamin C and dairy products, as well as physical fitness, appear to decrease the risk" and that is not an answer a doctor will be happy of giving out.

His conclusion was that it would have no effect whatsoever.

Exactly What American Med Schools Want (5, Insightful)

damn_registrars (1103043) | about a year and a half ago | (#41836857)

The American Medical Schools select for automatons who can memorize and regurgitate vast amounts of data. Talk to any physician who graduated from med school in the past 20 years about what they took as undergrads and they'll most likely tell you they don't remember the courses specifically because they memorized them to pass (and then promptly forgot them).

The real question is whether or not this is the best way to train our future health care professionals. While indeed there are some really good physicians coming out of our med schools - and even some of those who memorized their way through undergrad will be great physicians - we have also excluded from selection many who would have been excellent caregivers based on their inability to memorize quite as quickly as their classmates.

lots of colleges put to much on craming for tests (2)

Joe_Dragon (2206452) | about a year and a half ago | (#41836995)

lots of colleges put to much on cramming for tests and not much on being able to do stuff in a real setting. Closed book / closed notes / no Google tests only test memorizing.

What professionals one who is real good at test cramming or on who knows most of the day to day stuff and how to lookup the other stuff that they need at the time they need it. is better?

Were I an instructor, I would test both. (0)

Anonymous Coward | about a year and a half ago | (#41837101)

On their own looking stuff up, any resource they desire, but on their own.

Re:Were I an instructor, I would test both. (1)

bbelt16ag (744938) | about a year and a half ago | (#41837317)

and they wonder why they cant find anybody to place in positions. between this and the ads to 10 years exp in such and such and if you 9 you are out BS.

Re:lots of colleges put to much on craming for tes (1)

Ambassador Kosh (18352) | about a year and a half ago | (#41838365)

In my engineering classes they are pretty much allowing us to use books, notes, homework, previous exams, and pretty advanced calculators like the nspire but the also give FAR more realistic problems.

The exams tend to be very difficult and sometimes have unrealistic time constraints (ie nobody in a class of 100 people or so finish the exam).
Pretty much the only thing we can't use is a laptop/cellphone etc or anything with an internet connection.

It would certainly be nice if I could use matlab, python or excel on exams.

Re:Exactly What American Med Schools Want (1)

epSos-de (2741969) | about a year and a half ago | (#41837481)

Yes, exactly. The learning of today seems to be more about memorizing than about actually understanding. I remember that during my university years, I met a woman who did pass the most difficult exam by memorizing answers from examples and previous sample questions from previous exams. The funny thing was that she had no time for studying and did the preparation in one week, but passed with the highest score of the semester and is now an HR manager at a major German company. She claimed that she did not understand half of what she wrote down during exam.

Re:Exactly What American Med Schools Want (1)

demonlapin (527802) | about a year and a half ago | (#41837729)

The American medical school system accepts all sorts of students, so long as they have high GPA's and MCAT scores. When you get to medical school, though, you realize why all the traditional premed students have been cramming from day one. In med school, the dumbest guy in the class is pretty sharp. Tests tend to focus on the most arcane material imaginable, and so you have to memorize trivia. Experience helps. Of course, trivia is what medicine is really all about - the common things are common, and so easily dealt with. The odd cases are the ones to learn how to recognize.

Re:Exactly What American Med Schools Want (0)

Anonymous Coward | about a year and a half ago | (#41838155)

In med school, the dumbest guy in the class is pretty sharp.

Sharp? I think it's more like good memory.

The only sharpness you need is a bit of sense so that you don't produce "Toronto" answers like Watson.

Re:Exactly What American Med Schools Want (1)

Kurofuneparry (1360993) | about a year and a half ago | (#41838589)

Fourth year medical student here.

The comments above are quite true and in no way exaggerated. Being a good physician after passing the USMLE or COMLEX is mostly coincidence. Then again ... I'm an idiot ....

Before surgery, you must update Java. (0)

Anonymous Coward | about a year and a half ago | (#41836989)

Or resubscribe to JAMA, something like that.

IBM's health plan will demand it (2)

gelfling (6534) | about a year and a half ago | (#41837095)

No humans involved. Only computers and the occasional tech support call to Bangalore.

Dang, you had my hopes up (1)

WillAffleckUW (858324) | about a year and a half ago | (#41837099)

When I saw the twitter feed headline, I was hoping Lucy Liu was going to attend UW Medical School.

Sigh.

Dr. Watson, you're needed here!

good and bad by Dr Schteve (0)

Anonymous Coward | about a year and a half ago | (#41837255)

Couple of big things from Medical School:

80% of diagnosis can be made from history (talking to the patient/family) alone. Getting even just the key points from a patient interview into a computer will be a HUGE pain. So many times, the key points for diagnosis are very subtle.

The vast majority of medical tests are fishing trips. They are there for either ass-covering or because you have no idea what is happening (then a consultant asks the patient a few questions and gives the diagnosis, before the test results come back..).

Any new tool/gadget/gimick will be utilised immediately, especially if the government/insurance will pay for it, especially if it is expensive.

Medical school (first 2-3years of training) tests rely on memorise, regurgitate answers, purge your memory to make room for the next exam. Watson will do GREAT at this.
CLINICAL school (last 2-3 years before graduation) relies on patient interaction and clinical judgement. Things like "look at them from the end of the bed and get an impression". Even having a "thoughts and feelings" input field, Watson is going to suck at this.

Watson may be a useful tool for sifting through research (especially if it is good for getting rid of the rubbish articles) or analysing ECG/CT/MRI (but there is still a lot of clinical context needed). But remember, it is a tool. nothing more, nothing less.

I don't want to see this implemented as "run them through all of the tests and get Watson to spit out the summary" when it is a clear clinical diagnosis. With the way that over-testing takes place, we all know that this route is nearly inevitable.

-Dr Schteve

/. needs a watson editor (2)

NikeHerc (694644) | about a year and a half ago | (#41837267)

"I.B.M's Watson is headed to the Cleavland Clinic Lerner College of Medicine of Case Western Reserve University...

And if /. can't afford its own watson, how about a spell checker??? They are much cheaper and much, much less likely to make mistakes.

Spelling check... please... (1)

debio (695867) | about a year and a half ago | (#41837645)

"I.B.M's Watson is headed to the Cleavland Clinic" Really? It's the _Cleveland_ Clinic. Seriously "samzenpus". That's pretty bad. Even worse for the fact that in the next sentence it's spelled correctly. A bit too much Halloween bubbly perhaps?

Virtual Doc: You've got: leprosy. (1)

Joe_Dragon (2206452) | about a year and a half ago | (#41837981)

Virtual Doc: You've got: leprosy.

Re:Virtual Doc: You've got: leprosy. (0)

Anonymous Coward | about a year and a half ago | (#41839131)

no, it's Lupus.

wait, it's never Lupus...

An Open Watson please! (1)

Slur (61510) | about a year and a half ago | (#41838043)

What I would love to see is Watson's training interface on the Internet, as a service, with anyone able to pick a domain and contribute expert knowledge, whether in the form of questions Watson should ask, answers to those questions, or even just links to sources of relevant information. Through a crowd-sourced approach Watson's capacities could be so much more quickly developed. By keeping each user or group in sandboxes and maintaining knowledge in each domain more or less separately, there would be no problem for those who just input nonsense, or wrong information, because Watson can build up a reliability profile and consider that in later recombination of its knowledge.

Of course first there has to be a nice easy way of making Watson nodes that can be widely deployed. Frankly this is one area of research that deserves all the money we can afford to throw at it. In the future Watson will be able to derive new hypotheses and new knowledge obtained from inference, and it will certainly accelerate our research capacity.

Re:An Open Watson please! (0)

Anonymous Coward | about a year and a half ago | (#41839235)

Hello, my name is bucket. What's your name?

EMR is the missing link (0)

Anonymous Coward | about a year and a half ago | (#41838473)

Watson won't work until a complete patient chart is available; what passes for them today is just marketing fluff. Then rules engines like Watson will become a fast way to access the chart but nothing more because they will never be able to really communicate with a patient.

Boy I am excited! (1)

hundredrabh (1531761) | about a year and a half ago | (#41838573)

Soon we will have affordable, quality healthcare for all.

Next goal world peace!

Re:Boy I am excited! (0)

Anonymous Coward | about a year and a half ago | (#41838903)

Next goal how to deal with rapid over population

Re:Boy I am excited! (0)

Anonymous Coward | about a year and a half ago | (#41839269)

Better healthcare results, statistically, in less children being born. If cost of doctors were the problem, cheap doctors would cause a initial jump in the population but then it'd grow slower than before. But I think that by "for all" he means in the developed world, where overpopulation isn't a problem (if it were it wouldn't be considered developed).

Re:Boy I am excited! (0)

Anonymous Coward | about a year and a half ago | (#41839243)

I think we need robo-surgeons and robo-drug-researchers before that will happen.

Re:Boy I am excited! (0)

Anonymous Coward | about a year and a half ago | (#41840583)

World peace is easy: just kill everybody.

Hospital Pay System (1)

bigtreeman (565428) | about a year and a half ago | (#41838785)

He might discover the complexities of designing a health department pay system, since IBM screwed up the Queensland system so badly.

EMH (0)

Anonymous Coward | about a year and a half ago | (#41839227)

"Please state the nature of the medical emergency"

Re:EMH (0)

Anonymous Coward | about a year and a half ago | (#41846535)

Dammit, Jim. That's what I was going to say.

Image of the beast (1)

3seas (184403) | about a year and a half ago | (#41840309)

Stone/earth is what computers are made of and they run images of human/beast thought processes...... So what is coming is ......

There is more to life than just abstraction, which is what language is. There is more to reality than what our abstraction identify...

Some things worth considering.

Check for New Comments
Slashdot Account

Need an Account?

Forgot your password?

Don't worry, we never post anything without your permission.

Submission Text Formatting Tips

We support a small subset of HTML, namely these tags:

  • b
  • i
  • p
  • br
  • a
  • ol
  • ul
  • li
  • dl
  • dt
  • dd
  • em
  • strong
  • tt
  • blockquote
  • div
  • quote
  • ecode

"ecode" can be used for code snippets, for example:

<ecode>    while(1) { do_something(); } </ecode>
Sign up for Slashdot Newsletters
Create a Slashdot Account

Loading...