Contract with VA to use thier 9 million active (as well as former) member medical history in regards to data on prostate cancer. There have been very limited studies on prostate as to cancer cause and follow up. Mainly because disease is for 50 and up (60 mostly). Those that get cured drift away. Others die off. There are only 3 stnd methods of treatment all poor. There is zero concensus after a 100 years. Some believe its low testosterone as one ages. Others believe its testosterone itself that cause prostate cancer. Others believe its DHT converted testosterone. Others believe its first a low testosterone and then the testosterone is increased. The VA give blood tests called labs regularly to all 9 million members several times a year. They should include as standard labs for PSA marker, testosterone, DHT, estrogene and lutinizing. Shortly if not already available Watson could generate data on 9 million plus millions that have passed on to determine which of the named hormones/issues contribute directly to diagnosed prostate cancer. In my case, had a pituitary benigh tumor where testosterone dropped to nothing. After removal, testosterone back in normal range. Then PSA shot up dramatically, then leveled off. Then pituitary had issues again with testosterone low. Took prescribed testosterone. Then/since PSA keeps increasing dramatically. Biopsy indicated cancer. I have 5 years of labs for PSA, testosterone, DHT. I think its pathetic now that there is cloud computing and easy mgnt of data that using the VA a perfect captive market analysizing data each of these markers/hormones. VA captures weight every visit and other particulars. Those using the VA do not fade away once cured or even with slow growth cancer so data is massive as nearly all VA patients are older than 50. Donate this effort for humanitarian and get tax write off. Related idea for health. Set up a self serve as well as comprehensive data base for all diseases. Most doctors are not nearly up to date in the myriad of data or studies on all the diseases in their field. Develop 2 apps. One for patients, Show image of front/back and sides. Let them click on part of body they have pain/problem with. Then have Watson (after web crawling world data base of all diseases) display all the symptons for this disease. Let patient check off those they think they have. Next, let them list all other existing issues just like every stupid form you have to fill out over and over and over again at every doctor, dentist etc.Then IBM could sell this app to medical health companies to funnel traffic to them. Same type of version could be used for a doctor during/after a visit. Dr could enter by voice/transcribe what patient said. Watson could come back after reviewing all the diseases in the world against the listed symptoms and make determination of what disease(s) it is and recommend testing/treatment and most recents trials/studies. I dont want some doc trying to keep up with all the health/medical information being developed. Actually, most first line doctors could be eliminated or at least first vists be more productive. Medical health could reach around the world just like mobile phone. This will make people take ownership of their health- dis/ease. This is what is recommened before you visit doctor anyway. List all symptoms, pain, questions and related to be discussed at first meeting. Dont sit their like a log and have him guess which of the 100s of thousands of diseases you have. I was the one reading my labs with the PSA going up over the years not any of the doctors assigned to me. I
See this idea on ideas.ibm.com
I gave all that info in the idea above. Cull the VA for prostate cancer data via the blood labs on 9 million vets that use the VA health care service until they die so that trials and studies can be continuous and on going. Benefit is pro bono tax deduction.
Second app for medical heath care companies to assembly line first stage health care world wide. A doctor in India could treat me for an ear infection for 2 bucks rather than a $350 office visit to a general practitioner. Its so last century to expect some doctor to synthesize the patients symptoms for all the diseases that portray them. Web MD gives a limited canned oh you have this. Why not have a dynamic app that runs on Watson that can search the entire world disease data base for initial diagnosis. I have read where the inside stats on doctor mistakes, mis diagnosis and under diagnosed is between 25-40%.
Then people could also get their DNA mapped and tied to the app so Watson could also search their DNA and the markers identified for the symptoms being reported.
This is certainly a noble idea, unfortunately this ideas board is specifically for feature requests in the Watson Assistant service. Someone could implement this use case using the service, but the service itself is not the right place for this.
There is an entire org under Watson Health tackling difficult use cases, maybe this is something they could think about