How many steps/repetitions for gait and motion analysis

[Tomer Yona]

Hey everyone,

I wonder if someone has a reference or reasoning to choose how many repetitions of each motion is required for motion analysis?
Additionally, if using an IMU system for gait analysis and not constrained by lab size or force plates, will you think it is appropriate just to use a longer walking distance and fewer repetitions?

I'll be happy to hear your thoughts!

Thanks, and have a great week,
Tomer

[Samuel Acuna]

Well, as with most research, it truly depends on the research question you are asking.

For example, if you are looking at average cyclic data (e.g., gait) you will need to collect multiple repetitions. But if you are trying to examine specific behavior during gait, you might opt to extract a single representative stride of the data for analysis because the average will lose some of the details that might be important.

In general, I think as long as your approach is justified you will probably be fine. You can cite a paper that did something similar and used a certain number of repetitions, or even better if you can find a paper that recommends minimum collection times to get good results. Also, you can justify your repetitions based on your limitations (e.g., if a patient cannot perform a motion too many times). You just have to interpret your results with respect to the limitations of the study.

As far as the tradeoff between longer walking distance and multiple repetitions, again it depends on the research question. Likely, It depends on how many steps you need to acquire to answer your research questions. I would just do the longest walk (without inducing fatigue) you can manage, and then repeat it the minimum number of times to get the necessary number of steps for your analysis.

[Edmund Cramp]

The traditional view was to set a total walkway length of three to six metres, but the actual length depended on the gait lab size - normally you would collect data for the entire length of the walkway if possible but only perform an analysis on the data from the center of the walkway. This means that you are looking at the "normal" gait and not the subject starting to walk or slowing down at the end of the walk. In the clinical world the collected data would often be averaged but once the average gait was seen, then the actual analysis would be based on the raw gait cycles that were relatively well matched to the average gait data

I would strongly recommend reading these books which document the gait analysis environment and methods:

• Gait Analysis: Normal and Pathological Function Second Edition. Jacquelin Perry, Judith M. Burnfield.
• The Development of Mature Walking. David H. Sutherland, Richard A. Olshen, Edmund N. Biden, Marilynn P. Wyatt.
• Clinical Gait Analysis Theory and Practice. Chris Kirtley.

An issue with the IMU data environment that you mention would be that it's going to be difficult to verify the subjects' walking environment, did they step to one side slightly, lengthen their step to avoid a crack in the pavement or a piece of paper on the floor? Are they changing their gait slightly because the path is not a perfectly straight and level line? In addition, looking the the raw force plate data and including it in any analysis is very helpful.