What I find most interesting is the possible false detection rates that widescale deployments of thermographic systems may well encounter. I would suspect that in light of the unknown accuracy of the majority of new systems being deployed, their false positive rate, when compared with Meditherm's would likely be higher.
their false positive rate, when compared with Meditherm's would likely be higher
Or it could be lower and their false negatives (people with fevers they miss) might be way higher. Hard to know without ground truth (i.e., doing a clinical reading of everyone who is scanned by the thermal system).
Very true John and not something to discount. I was looking at it more from the angle of: detect a fever (wrongly in the majority of cases) -> trigger a response (of some sort). What resources will be needed for the responses?
Rather than an absolute false negatives/false positives test. Much like FR deployments where the second is all that is considered since the first is so hard to measure.
Interesting also was that Meditherm's data comes from decades of field use - - but given their normal market, it was likely collected primarily from within a clinical environment, which is a much more controlled environment than is practical to arrange for mass security screening purposes.
By the time a patient is actually being engaged by a health professional in a clinical environment, they've usually been idle in a controlled waiting space for 10-20 minutes or more.
So many unknowns in this application and it's validity against expectations.
IPVM's doing a very important service in focusing on this segment. Much appreciated!