# Study: passive detection of Atrial Fibrillation using smartwatch
**URL:** https://forum.fudiabetes.org/t/study-passive-detection-of-atrial-fibrillation-using-smartwatch/3762
**Category:** Tips, Tech & Gear
**Created:** [March 21, 2018, 10:17pm UTC](https://forum.fudiabetes.org/t/study-passive-detection-of-atrial-fibrillation-using-smartwatch/3762 "2018-03-21T22:17:23Z")
**Posts on this page:** 3
**Page:** 1
### Author:  [@Michel](https://forum.fudiabetes.org/u/Michel)
#### Post date: [March 21, 2018, 10:17pm UTC](https://forum.fudiabetes.org/t/study-passive-detection-of-atrial-fibrillation-using-smartwatch/3762/1 "2018-03-21T22:17:23Z")
Here is another study that shows passive measurement and detection of a medical condition using a smartwatch (an Apple watch in this case), in conjunction with a neural network (AI/ machine learning):
> **[Passive Detection of Atrial Fibrillation Using a Commercially Available...](https://jamanetwork.com/journals/jamacardiology/article-abstract/2675364/)**
>
> This study aims to develop and validate a deep neural network to detect atrial fibrillation using smartwatch data.
The detection rate was an outstanding 97%. This diagnosis tool is not available commercially (yet). This comes only a few days after another study showed a way to passively measure potassium blood levels using another smartwatch sensor:
> [@New Apple wristband sensor can detect potassium blood levels](https://forum.fudiabetes.org/t/new-apple-wristband-sensor-can-detect-potassium-blood-levels/3690):
>
> Now that’s pretty cool: The AliveCor KardiaBand, a sensor compatible with the Apple Watch, can detect dangerous levels of potassium in blood with 94 percent accuracy. […] [It] is a sensor that snaps into a slot on the watchband. The user touches the sensor, which then takes a reading of the electrical activity of the heart, called an electrocardiogram (EKG). This reading can reveal abnormal heart rhythm and atrial fibrillation (AFib), and the sensor sends the information to an app. Yester…
For us, of course, the holy grail would be remote sensing of blood glucose without a skin-embedded sensor. It may not be so far after all. Another 10, 15 years?
---
### Author:  [@Boerenkool](https://forum.fudiabetes.org/u/Boerenkool)
#### Post date: [March 22, 2018, 6:40pm UTC](https://forum.fudiabetes.org/t/study-passive-detection-of-atrial-fibrillation-using-smartwatch/3762/2 "2018-03-22T18:40:23Z")
> [@Michel](#):
>
> For us, of course, the holy grail would be remote sensing of blood glucose without a skin-embedded sensor. It may not be so far after all. Another 10, 15 years?
I hope so. I just find it very hard to believe that it would be possible to measure blood glucose accurately with a non-invasive sensor.
---
### Author:  [@Chris](https://forum.fudiabetes.org/u/Chris)
#### Post date: [March 23, 2018, 6:25pm UTC](https://forum.fudiabetes.org/t/study-passive-detection-of-atrial-fibrillation-using-smartwatch/3762/3 "2018-03-23T18:25:36Z")
It is possible today, just currently very expensive in terms of the processing power required to make the measurement. Over the years a few companies have made a finger sensor (like an O2 sensor) that would give an accurate glucose measurement. They have been successfully made, they just aren’t priced low enough to be commercially viable. As processing power increases, this will be routine.
If you are a science-person, the reason is that the glucose absorption in the water window that our bodies give, is very small (second overtone I believe) as opposed to oxygen which is a primary absorbance in the water window.