That's an excellent question, and it actually has a very big application for the public sector to think about. To give you an updated version of that, we are now very heavily invested in oncology in this cognitive computing we call Watson. I was with one of the federal ministers in our labs in New York—unfortunately, it was in New York, but we'll have it in Canada soon—where we were looking at an oncology patient. What it does is it looks at thousands and thousands of medical records, including the individual's, but also all applicable cases, and it asks what the best outcome would be for the patient. They think about the patient when they do it.
It has a ranking of diagnoses that you could use. The doctor does the diagnosis in the end, but the interesting thing is that the third best diagnosis and treatment, which was the most conservative, was full radiation and full chemo. That was the third best according to Watson, 55%. There was a 90% and an 85%. With hybrid versions where the patient does not lose her hair—the patient was a young mother—she has better outcomes. Actually, the best part is that, at the end it cites out of all those millions of medical records that they're looking at the exact trusted cases to explain why it says this.
The doctor can decide whether that's what they want or not. By the way, it does it in three seconds. It doesn't do it in years. It does it very, very quickly and prevents many other invasive tests of other types that patients generally have to deal with.
I'll give you one other example for disruptive purposes. We put it out in a competition to a bunch of universities. U of T came in second in the competition about two months ago, and they actually applied it to the legal system. The legal system wasn't what we were thinking about, so this is an example of where we said to put it out in front of people and see what people will do. They took 20 years of family case law, put it out, and asked Watson to look at the case against all of the 20 years of Ontario family case law and come up with the right precedents for that case. Of course, doing it very, very quickly, it had five relevant cases, the killer cases, where now the lawyer has to convince a judge.
The interesting thing is that it now disrupts the whole legal system. It's not a thousand-person law firm that wins; a two-person law firm or a one-person law firm can actually get the same information. By the way, it takes months and months and months out of that legal system, if you will, and all of the discovery and so on. You can think about this for policy use and other things that we can talk about at length.
I'll just come back to your point. We're very proud of it because it is something that is there, but when it sits in a back room, it's not useful. Bringing it out and actually having it apply to things that are really important, that's what we're up to. Canada can lead, maybe because of our footprint, but actually because we have a point of view to build this in this collaborative way.