Attention all in the Vort collective: I hope you all don't mind if I take a few bytes of bandwidth to request some help with the R&D I've been working on... which is noninvasive blood glucose measurement using RF/microwaves. The attached pic shows the results for just one of the diabetics tested; for this one we could get a good calibration on 82 data points (taken in Feb 2010), and then the calibrated equation accurately estimated the remaining 120 samples which were taken thru March. Follow-up testing in June also gave good results with little degradation. Predictive accuracy over time is a major accomplishment in this work.
We have a database of ~87GB, most of which was on five Type-1 diabetics over the course of 2 months; clinical lab-grade blood chemistries for most of that data. During RF scans we are also taking skin temperature every 100 millisecs... Our investor has given us until the end of the year to improve our calibration/predictive algorithms as much as possible before we market the technology for the next phase of development. We are currently at +-20% accuracy for ~80% of our samples (~1000 samples on the 5 test subjects). The technology is not optimized, so this may be all we can hope for with the current sensor design and algorithms. But, we need to use the time left to make whatever improvements we can... I am in search of some very bright individuals with expertise in mathematical modeling and bioelectromagnetics; perhaps statistics, but targeted toward medical device testing. Knowledge of RF Scattering Parameters (S-Params) which come out of a modern Network Analyzer (Agilent PNA-5230) would also be very helpful. We already have some very extensive MatLab code which builds mathematical models, one term at a time, and it may be better to add to this rather than creating from scratch. IF you're very competent and like a real challenge, and want a break from the E-Cat fiasco, then please contact me @: [email protected] or [email protected] There are now 366 million diabetics in the world, and they have been in need of a truly painless way to measure their blood sugar. You could be one of the keys to solving the challenges which make this a reality for them... Thanks for your time... Now back to your regularly scheduled E-Cat frustration! :-) -Mark Iverson
<<attachment: TS11_CalPred.jpg>>
