In several studies which aim is to test or evaluate static or dynamic balance, is usual that the measuring tool are force plates, platforms or pressure insoles. In order to perform the balance recording, the researchers have to set the appropiate settings, which includes the sampling rate. Some studies had reported sampling rates ranging from 200 to 1000 Hz or more, depending on the tool.
Having in count that each one of the main sensory systems involved is supose to react and modulate balance at very low frequencies (compared to the sampling rate that the measuring tools usually have in most of the studies), is it really necessary to set the sampling rate to high?
The Nyquist sampling theorem dictates your sampling frequency. You first must determine the highest frequency intrinsic to the process you are trying to record. The Nyquist theorem states that your sampling frequency must be at least twice that of the highest intrinsic frequency. If you don't, aliasing will occur. Often, the ideal sampling frequency may be more like ten times that of the highest intrinsic frequency.
Good question, and the answer depends on what you want to analyze: if you want to do a frequency analysis or measure fast balance tasks, you might need a higher frequency (200Hz?); if you're only interested in sway path length or standard ellipse are, it's better if it's not too high (about 50 Hz), because noise can enhance sway path length. It depends also on the quality of your force plate (noise, spatial resulution of the CoF). I would suggest to a) put a weight on the force plate to record the noise and b) have someone perform the balance task that you're interested in and look at the frequency spectrum and CoF path length. You can start with a high sampling rate and downsample to see what changes.
I agree with Andreas' recommendation. For most of our postural control investigations, we use 50 Hz but the sample rate will increase based on the task and if we sync the force plate with another instrument. The force plate typically needs to be a multiple of the sample rate of the other instrument, so that is likely why you see rates as high as 1000Hz.
The Nyquist sampling theorem dictates your sampling frequency. You first must determine the highest frequency intrinsic to the process you are trying to record. The Nyquist theorem states that your sampling frequency must be at least twice that of the highest intrinsic frequency. If you don't, aliasing will occur. Often, the ideal sampling frequency may be more like ten times that of the highest intrinsic frequency.
A word of caution - all outcome measures vary with sampling frequency. You may want to consider this if you plan on using the protocol for more than one experiment.
Thank you very much, to all of you, for taking your time to answer this question ... i also would like to know if there's any paper related to the influence of different sampling rates on the same balance testing
Force platforms should not be used to infer balance because sway measures relate poorly to risk of falling.
stevenrobbinsmd.com
Hi Javier,
I very much agree with and recommend to follow Andreas' and Gordon's comments.
With respect to "fast balance tasks" (cf. Andreas' comment), like e.g. swinging the bat and hitting a baseball, you might want to calculate the respective braking and acceration impulses from CoF-data, and even calculate higher drivatives of the space-time funciton (i.e. velocity-time function, accelaration-time function, and even jerk). Then you'd need higher sampling rates than 200 or even 500 Hz.
In our biomechanics lab using a Kistler force plates, we therefore use a standard sampling rate of 1000 Hz.
Higher sampling rates are also recommended when assessing reactive balancing behavior in at-risk of falling subjects, when exposing them tocritical situations (e.g. mimicking slips on slippery ground by jerking the platform foreward or backwards, etc.) or to secondary task (i.e. cognitive load).
Regards,
Klaus Blischke
Hi Javier,
This article may be helpful to you. It looks at both sampling frequency and filtering techniques in relation to static balance control.
Cheers,
Rebecca Robins
- Badges
- Science topic