GGIR for waist worn assessment

[Dylan]

Hi everyone.

Really great package and platform. 

I understand that the current GGIR package isn't optimised for waist worn accelerometers.  

Is anyone actively working on looking to optimise for waist-worn  measurement?

I would be interested in collaborating.

Best wishes, 

Dylan 

[Vincent van Hees]

Hi Dylan,

Thanks for your message and sorry for the slow reply. I am currently working on a two projects that relate to your question.

You are correct that many of the default GGIR argument values are tailored to wrist-worn accelerometer. However, GGIR is designed to be adaptable to any sensor location. There are about half a dozen of publications that have proposed cut-points for hip or waist worn accelerometers to study physical activity, you can use those cut-points in GGIR. So, I think a basic physical activity analysis is already possible.

Less well understood is the value of GGIR's sleep detection for waist-worn accelerometers. If you know of any polysomnography data sets where an accelerometer was worn on the wrist then that would be great. Also, datasets with sleep diary and hip-worn accelerometer could be valuable.

Let me know if you want to chat, and we we can explore how you best contribute.

Best wishes,

Vincent

Dr. Vincent van Hees

Independent consultant

https://accelting.com/

‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐

--
You received this message because you are subscribed to the Google Groups "R package GGIR" group.
To unsubscribe from this group and stop receiving emails from it, send an email to RpackageGGIR...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/RpackageGGIR/5f0a4fce-4e1c-492a-95b5-56379249fc5an%40googlegroups.com.

[Dylan]

Thanks, Vincent, 

Are there any examples of code in GGIR that would optimise output for waist worn sensors?  I can see there are example shells  for the wrist worn

Best wishes, 

Dylan

[Vincent van Hees]

Hi Dylan,

Good question.

Most of the GGIR functionality should be wear location agnostic, with a few exceptions:

1. Physical activity intensity cut-points: These should be tailored to body location (see vignette for links to literature).

2. Sleep: The current default are specific to wrist location. If you specify HASPT.algo = "HorAngle" and sensor.location = "hip" GGIR attempts to identify the main lying down period during the day based on sensor orientation. It says "hip" but this should also work for "waist". Note that there is no literature at the moment to support the validity of this, it is just an heuristic method that uses common sensor about the orientation of people when asleep. If you are in a position to run a validation study to help validate this then let me know.

I agree that a more elaborate tutorial with examples tailored to studies with hip/waist worn sensors would be valuable. Although, I think it may be better to prioritise efforts to evaluate and where necessary improve the functionality first.

With kind regards,

Vincent

Dr. Vincent van Hees | Independent consultant | https://accelting.com/

------- Original Message -------

To view this discussion on the web, visit https://groups.google.com/d/msgid/RpackageGGIR/33a7bcae-68bb-42a8-984f-759420c74d7en%40googlegroups.com.