I want to know in your opinion which is the minimal sampling frequency required and the best filter to use. Moreover which values do you take into account: I mean the ones from the vertical axes of the device (aligned with the tibial bone) or the sum of the three acceleration recorded ( sqrt (x^2 + y^2 + z^2))?
Acceleration is usually sampled at 200Hz unless you're coupling with other biomeasures in which case you sync all with the highest sampling rate and downsample offline. I recommend using resultant data and converting to either m/s/s or g-force.
Thanks Dr Josephson,
I use IMUs alone for clinical practice.
What type of filter do you use for data smoothing?
Butterworth has less of a "ripple" effect or a root mean square moving average window. I recommend checking other publications to see what they did, then play with the filtering yourself a bit to determine what works best for your research and why.
- Badges
- Science topic
- Similar topics
- Biological Science
- Kinesiology
- Running