Description

author: M.Millard date : 20/1/2020 - 11/5/2022 email : matthew.millard@inspo.uni-stuttgart.de

This repository contains the collection of scripts (without data) used to process the sit-to-stand dynamic balance of younger and older adults as described in Sloot LH, Millard M, Werner C, Mombaur K. Slow but Steady: Similar Sit-to-Stand balance at Seat-Off in older vs. younger adults. Frontiers in sports and active living. 2020:144.

Overview

This is intended as a code example to help other researchers apply technical measures to analyze the dynamic balance of people. Here the balance measures considered are: the foot-placement estimator, the capture-point, and the functional base of support as described in Sloot et al. In this repostory you will find the following pipeline to analyze dynamic balance data:

1. Kinematic data: c3d files and Visual3D data. The Visual3D data includes: the center-of-mass (COM) position, velocity, linear momentum, angular momentum, and the whole body moment of inertia about the COM. This data (not included in the repository) is stored in /inputData and loaded into a generic structure by making use of the configuration (e.g. configE01.m) files and the configuration scripts (procesConfig.m and processConfigCommon.m)

2. The collection of scripts in code/matlab/ performs 4 functions:

• main_ProcessSubjectFunctionalBOSAnalysis.m : evalutes the generic scalable base-of-support model.
• main_ProcessBalanceData.m : uses the kinematic data to evaluate the FPE, CAP, and functional base of support
• main_StatisticalAnalysisPreprocessing.m : extracts measures at specific times of interest and packages these measures in a structure for later statistical analysis
• main_PlotFrontiersSpecialIssueFiguresR2.m : generates the figures that appear in Sloot et al.

Any file in code/matlab/ that begins with main_ is designed to be run manually. Not all main_ files are fully independent since the data in Sloot et al. is processed in the following stages: main_ProcessSubjectFunctionalBOSAnalysis.m, main_ProcessBalanceData.m, main_StatisticalAnalysisPreprocessing.m, and main_PlotFrontiersSpecialIssueFiguresR2.m.

3. To get started in applying these technical analysis to your own work, I recommend starting with the main function that does much of the number crunching used in Sloot et al.:

• main_ProcessBalanceData.m
  • Preprocessing
    • getC3DTrialData (line 282)
    • getAnthropometryData (line 298)
    • getWholeBodyTrialData (line 318)
  • Technical analysis
    • constructFootFrames (line 411)
    • process3DFootPlacementEstimator (line 446)
    • processCapturePoint (line 470)
  • Metric evaluation
    • processDistanceToConvexHull (line 542)
    • processDistanceToConvexHullUsingNormBosModel (line 548)
  • Movement segmentation
    • segmentDataKMeans (line 664)
    • extractSitToStandSequence (line 702)
  • Generation of RBDL model (requires a special version of Manish Sreenivasa's ModelFactory)
    • writeModelFactoryModelFile (line 878)
    • writeModelFactoryEnvironmentFile (line 897)
    • createModel (line 908)

Each of these functions appear in code/matlab/ should be read with the Sloot et al. paper at hand. Some functions, particularly the technical analysis functions, are quite detailed and will take patience and time to follow.