Our BANN is Simple Really

Trying to explain innovations in medical science has to be difficult, for most people, most of the time. When that science is combined with advanced statistical theory and computer technology the problem gets a lot bigger, fast. The word about our research is getting out, which is good, resulting in lots of smart people wanting to understand more. Here’s the simplest explanation we’ve been able to come up with.

Our technology is taking readings output from patient monitors and matching those with data we have in the history bank. When matches occur it checks the progression of the historic readings and records subsequent measures in the progression. When it finds enough matches between the now readings and what happened subsequently in the history bank it determines similarities with the current data set. When there’s sufficient correlation between the historic and current data sets it’s able to suggest what might happen in the future, based upon what happened enough times in the past, in the same patterns. Ultimately, with enough matches the software can output indications of what’s likely to happen in the future.

Statisticians might describe this as a prediction with a specified probability.
Clinicians would explain they’ve seen this pattern before, and based on past experience, predict what might happen in the future.
The layman would point to the barometer, knowing when the pressure falls rain won’t be far away.

It’s a simple as watching what’s happening now, comparing that with what happened in the past, and calculating the likelihood of that re-occurring. We’re comparing what’s going on now with previous models and deciding matches offer indications of what to expect in the future.

Our BANN is doing precisely that, and only that. Just what we would do, only a million times each minute.

It’s simple really. Just very hard to do!