Many research studies have incorporated the use of wearable devices into the study to measure the activity of humans in response to a variety of interventions. There are grossly three parts to that process:
1. Selection and adoption of the activity monitor – accelerometer or gyroscope based devices
2. Process for collection of the data – supervised activity or free living activity of the subject
3. Analysis of the data
Here is one tool for use in part3: on the analysis of the data:
When analyzing large data sets from activity data that is collected from commercial devices like fitbit or clinical devices, it is always tedious to analyze the data and draw meaningful information. Most of the tools are very customized by the companies that manufacture the wearable device.
However, a few python based tools are available that enable the analysis of the data into epochs as well as graphical representation of the data. One interesting one from UK has been used in multiple studies for GWAS analysis to correlate genotype to the phenotype such as the two below.
Doherty A, Smith-Bryne K, Ferreira T, et al. (2018) GWAS identifies 14 loci for device-measured physical activity and sleep duration. Nature Communications. 9(1):5257
Willetts M, Hollowell S, Aslett L, Holmes C, Doherty A. (2018) Statistical machine learning of sleep and physical activity phenotypes from sensor data in 96,220 UK Biobank participants. Scientific Reports. 8(1):7961
And rather than complex behavior patterns that can be confusing, their code generates summaries that address issues important to wearable analysis: sedentary behavior, physical activity and sleep.
It is available on github and if you have working knowledge of Python you can try it with your data or public data available from Open Humans.