Aggregated parameters question

[wengwongg]

Hi guys!

I'm just currently working on a dissertation project where I'm extracting gait parameters for MS patients. I therefore plan to use the preconfigured impaired pipeline. I'm just however confused about the aggregated parameters that are returned. I don't really have any experience with gait analysis, but is there a reason for the inconsistency in the returned aggregated parameters?

For example, there is average and variance for cadence and stride duration over all WB's, but not for gait speed? Also different aggregates are returned for different time intervals, e.g. WB 10 to 30 seconds only has average walking speed and stride length, which is not the case for parameters over all WB's?

I appreciate all the help I can get! :)

[AKuederle]

Hi, These are the parameters that we successfully validated with regards to clinically reliability and consistency. Paper on that will hopefully be published soon. Depending on the type of research you are doing, you can also define your own aggregations. Arne

…

On Fri, 7 Feb 2025, 16:21 Weng Wong (George) Hum, ***@***.***> wrote: Hi guys! I'm just currently working on a dissertation project where I'm extracting gait parameters for MS patients. I therefore plan to use the preconfigured impaired pipeline. I'm just however confused about the aggregated parameters that are returned. I don't really have any experience with gait analysis, but is there a reason for the inconsistency in the returned aggregated parameters? For example, there is average and variance for cadence and stride duration over all WB's, but not for gait speed? Also different aggregates are returned for different time intervals, e.g. WB 10 to 30 seconds only has average walking speed and stride length, which is not the case for parameters over all WB's? I appreciate all the help I can get! :) — Reply to this email directly, view it on GitHub <#206>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACSHBO4OULOJE6JR6C7W5BT2OTFPLAVCNFSM6AAAAABWWGK3LSVHI2DSMVQWIX3LMV43ERDJONRXK43TNFXW4OZXHEZDQNBZGM> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[wengwongg]

I see, thank you! 👍

[AKuederle]

Just as a side note on this. If you want to use custom aggregation, you could create a new Aggregator subclass, but it it likely much easier to just take the per_wb_parameters_ output of the pipeline and then apply your own aggregations directly. Though, this will skip the applying of the per_wb_parameter_mask_ which tells you which parameters are outside of expected ranges. Depending on what you are doing, it might make sense to consider this. You can check the code within the MobiliseAggregator (

mobgap/mobgap/aggregation/_mobilised_aggregator.py

Line 361 in 01bfd5e

# We remove all individual elements from the data that are flagged as implausible in the data mask.

) to see how we handle this.