There has been a market explosion of smart wearable sensors. The sales are predicted to raise considerably in the next 5 years and several large companies are adopting the technologies to further develop the products. The New York Times has published several pieces in their Science section pointing out the inaccuracies of the devices. Yet it is thought that these gadgets will disrupt clinical research, see ongoing discussion here
https://www.linkedin.com/groups/Will-health-data-collection-platforms-4322249.S.5877605859869102083?view=&item=5877605859869102083&type=member&gid=4322249&trk=eml-b2_anet_digest-null-5-null&fromEmail=fromEmail&ut=0gEk-1fPeNt6g1
I think that using proper statistical algorithms in these devices will bring us closer to truly personalized medicine. Yet as it stands today they do not seem accurate enough and seem to be using inadequate statistics for individualized assessment. I think that the Researchgate community could actually make a difference and open a platform for data standardization, exchange and analyses with the potential to change the diagnosis, tracking and treatment methods for various disorders of the nervous systems.
Well, I guess it is a balance between cost and accuracy. I guess you mean the sensors that monitor physical activity levels. They are relatively inexpensive in comparison to exercise equipment. If you want better more precise devices, that are accurate to the percent then cost is going to explode and no one will buy these, or the average person won't. In all honesty, I think cheaper is better in terms of just physical activity monitors as to get the masses moving is more important at the expense of accuracy. I have no problems exercising without these devices and at times get annoyed by all the gadgets around the place. It is all a matter of preference I guess. I don't know that I've really answered your question as maybe you are referring to other sensors. I have no idea what algorithms they use.
Thanks Deborah. I was referring to the algorithms rather than the sensor type. But the question of what components they use to build the sensors is interesting as well. I'm curious about the sampling frequency of the sensors too and how the algorithms decide on the counting, binning and estimation of parameters. The few interfaces that I have seen give bars and some give curves, etc. but the motion data (for example coming from accelerometers) is very rich in information. It would be nice to know what they focus on. Thanks again
I think this paper is a good example for mobile health.
"Outpatient Glycemic Control with a Bionic Pancreas in Type 1 Diabetes"
http://www.nejm.org/doi/full/10.1056/NEJMoa1314474
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis