Incident Detection Improvements

Hi all,

I'm looking for incident detection improvements when HR during activities falls below a minimum danger threshold.

That way, if you are in Cardiac Arrest during activities, incident detection automatically send your location to emergency contacts...

Thank you very much

  • This isn't Connect IQ relates, and you may want to submit the idea to Garmin:

    https://www.garmin.com/en-US/forms/ideas/

  • I know there is a patent over this type of biometric based trigger and alert notification system (US9547972B2) which might explain why it has not been implemented yet.

  • I deal with false alarms from physiological monitors all the time in the hospital. Most (90-99+%) physiological alarms are non-actionable in a healthcare setting with mostly stationary patients, who are actually sick, and hooked up to advanced monitors that give higher quality data than a watch. Optical HR is one of the worst offenders when it is used (EKG preferred over it if available).

    You have to consider what the acceptable false alarm rate is going to be for a user before they get annoyed and turn it off... and the first time they miss a false alarm and don't cancel it in time before it messages their contacts, they'll likely disable it. So, you need good specificity (low Type I Error) to not annoy your users, but you are going to trade sensitivity (increase type II error) and response time (longer delays, less sensitive thresholds, and longer averaging times with quality of data considerations) to the point where such a feature may become useless.

    tldr; that isn't what this sensor was designed for and it is doubtful that coding can adequately compensate to create a useful feature.