Modern medicine does wonders when it comes to targeted healing of our specific ailments - especially when it comes to mechanical malfunction of a body part. Surgeons can go to very precise organs and fix or even replace them with bionic parts.
While the health care system trains doctor to test, scan and treat targeted ailments, every now and then, we come across anecdotes of physicians not being able to think holistically to produce the correct diagnosis. Today's diagnosis is often done by individual symptoms or test results - not by connecting the dots of myriad other discrete symptoms or past health history of the patient. In other words, even the best medical technology is still ways off from treating the human body as a super-complex system.
A friend of mine has been seeing doctors for years with many individual but apparently unrelated issues - backache, fatigue, recurring eye redness and on and on. Every time she received "adequate" treatments for the specific conditions but they all kept coming back. Not until she, at the behest of another friend with similar history, researched and consulted multiple experts could she find her real underlying degenerative chronic auto-immune condition.
So what have big data and artificial intelligence got to do with this? Could be a lot.
In an engineering conference for the automotive industry (where I work), Rob High, CTO of IBM Watson Solutions, made a case for 'big data' and cognitive computing in solving complex problems including medical diagnostics. Integrating myriad medical data and past history in to a cohesive medical diagnosis can be a computing challenge. Even if we ignore the lack of holistic approach, modern health care in many ways is becoming victim of its own rapid progress. According to Dr. High, new medical information and discoveries now double every two years. Nearly, 80000 pages of new medical information are created daily. Yet, a typical medical practitioner hardly keeps in touch with a meager fraction of these new discoveries. An average doctor, processes no more than 5% of this growing knowledge base and is often miserably outdated. 81% of physicians don't even spare five hours per month to keep up.
In 2011, IBM demonstrated its own progress in artificial intelligence, data analytics and cognitive computing when its Watson computer program bested Ken Jennings and Brad Rutter - the all time best Jeopardy! champions in a three-day competition. What Watson demonstrated was its capability to understand the clues and contexts hidden in an 'answer' and then to analyze all relevant information and prior learning from its very large memory and produce the most statistically relevant outcome. Clearly the same technology has potential to lend a hand to develop more integrated medical diagnoses.
I am encouraged that Cleveland Clinic is looking to apply this technology for its expert "diagnosticassistant'. I also hope that someday, smart evidence-based (if not fully holistic) treatment as indicated by this Youtube demo about oncology treatment may become routine.