Computers as Doctors
The history of computer-assisted diagnostics is long and rich. In the 1970s, researchers at the University of Pittsburgh developed software to diagnose complex problems in general internal medicine; the project eventually resulted in a commercial program called Quick Medical Reference. Since the 1980s, Massachusetts General Hospital has been developing and refining DXplain, a program that provides a ranked list of clinical diagnoses from a set of symptoms and laboratory data.
And I.B.M., on the heels of its triumph last year with Watson, the Jeopardy-playing computer, is working on Watson for Healthcare.
Katie Hafner in the New York Times.
Wow, a step beyond physician diagnoses over the phone.
This is a welcoming and necessary change within the health care industry, in terms of both efficiency and access.
It would be interesting to see the places where computers fall short versus human doctors and vice versa .. and then try to parse out “why.”
I have patients come in all the time who express that they just want to know what is wrong with them, and yet when taking their history it is clear that they already know, they just haven’t accepted it yet. It is my role to explain it to them – to find the language and the illustrations that will give them insight into their condition. My role is also to lay out what is known and what is not know and to chart a path towards rationally clarifying any remaining uncertainties. These are roles that Watson cannot fill.
This is an interesting development in healthcare. Im curious to see how well it works.
Does anyone know when the new iPhone will come out that features an app that can diagnose any illness you might have?
I found this article to be very interesting. Doctors will probably scoff at the notion. Yet, the blood chemistry and health metrics of millions of people matched with their diagnoses and health status over time could create a huge database for predicting disease. A computer algorithm could easily compute the probability of various diseases and request other data (i.e. tests) that could narrow possible diagnoses even farther. This type of algorithm could also predict future health metrics (blood pressure increasing 5 or 10 years hence) and likely date of death. It might help doctors motivate patients to better manage their blood pressure and watch their weight if the report clearly shows they have already reduced their longevity (irreversibly) by, say, 7 years and will reduce it even farther in the coming years if they don’t change behavior.
I liked the hospital’s diagnostic computer in the movie Idiocracy. When Luke Wilson’s character went to the emergency room, the doctor gave him three identical-looking probes and said to insert one probe in his ear, one in his mouth and the last probe in his rectum. Luke reluctantly complied…
The computer returned an error code, which caused the doctor to take back all the probes. He then said… “no, insert this one in your butt, this one in your mouth and this one in your ear.
This looks to be the major benefit of Watson.
“Watson then mines the patient data to find relevant facts about family history, current medications and other existing conditions. It combines this information with current findings from tests and instruments and then examines all available data sources to form hypotheses and test them. Watson can incorporate treatment guidelines, electronic medical record data, doctor’s and nurse’s notes, research, clinical studies, journal articles, and patient information into the data available for analysis.”