Skip to content

cliffordlab/OS_Actigraphy_Toolbox

Repository files navigation

Open Source Actigraphy Toolbox

Matlab code to analyze actigraphy data. Each of following methods is stored in their own folder inside this repository.

Preprocessing

  • Bandpass filters
  • Cole and Oakley methods for converting accelerometer signal to activity counts
  • Conversion to time-day matrix format which is required by Rest-Activity Metric calculation

Rest-Activity Metrics

For calculating Rest-Activity Metrics, use convert2Hours to get required time-day matrix data format. Then use preprocessRA to eliminate missing hours.

  • Interday stability
  • Intraday variability
  • Least active 5 hours
  • Most active 10 hours
  • testRestActivityFunctions: Testing if new RA functions work as described by the author
  • testMissingData: Trying to find optimal method to combat against missing data

Circadian Rhythm

  • Cosinor Rhymometry
  • υτ (v_tau), a measure of stability in rest-activity rhythm
  • Dichotomy Index (I<O)
  • Autocorrelation Coefficient (r24)

Sleep Detection

  • Sleep detection with Oakley method

Visualize Data

  • Double plot for visualizing actigraphy data over a couple of days

References

For references, please look at header of each code.