During 2007 or 2008 I read that there was software that could process MRIs to know which part of a persons body it was looking at. I suggest that technique treated as a database creates the opportunity to compare say the size of the livers of 10,000 people from existing datafiles, then see if liver size correlates with reduced susceptibility to cancer. a plausible hypothesis as the liver makes P450 enzymes that process xenobiotics, some of which are regulated as carcinogens.
As such this big database is a place to find genetic immunity or resistivity to diseases. It is possible that people with big pancreases are less likely to be functionally diabetic as they might make more enzymes. Thus people with a family history of diabetes could IVF or other technique to choose an embryo likely to be immune to any genetic susceptibilities to disease.
so really its just a database applied to this MRI system finder technology I read about. Although as a database it could also be explored to find numerically suggestive data to create valid hypothesis. like big liver goes with less cancer is kind of obvious, the database might find things like lung branching fractal number suggests immunity to asthma or something where the numbers lead part of the research. Or possibly complete length of arteries (divided with) heart size predicts cardiovascular well being, yet can be linked to a genetics of arterial length.
<funny> its mass throughput non brain phrenology at p<.001! </funny> It could also work. It also makes genetic medicine more attractive. as people could physically reduce cancer, heart disease, diabetes all with one database. also this voluntary medical technology benefits more than just one treated person, it benefit all their descendants.-- beanangel, Mar 20 2012 What if it turns out that the comparative size of individual organs have no correlation whatsoever to disease susceptibility? If you must assume things, try to do it near the end of an idea, rather than predicating the entire idea on a single unsupported assumption.-- Alterother, Mar 20 2012 well the thing is that the system searches to find such correlations at P<.01 or better. Then humans can decide if they have plausability or not. The idea is the DB software
Also it works pretty well, it is published that women get drunker from the same amount of alcohol mg/kg because of bodyfat percentage. If people didn't previously know that this software would detect it.
Do people with larger livers metabolize xenobiotics more rapidly, its likely yet I do not know. It is published that the very elderly make less p450 enzymes which strongly affects the length of time drugs reside at their systems, thus there is some strong preexisting data supporting that this db could actually figure out big livers reduce cancer.-- beanangel, Mar 20 2012 Is there a database of all MRI's ? obvious beneficial analysis/mining diagnostic possibilities.-- FlyingToaster, Mar 20 2012 // Is there a database of all MRI's ? //
Theoretically. With patients' consent, imaging and other information can be made available anywhere via various med records networks. Collating the data would be a legal and logistical nightmare.
It's not implausible, and I'll defer to those who know better than I, but I think this idea is sketchy.-- Alterother, Mar 20 2012 random, halfbakery