A Data-driven Methodology to Characterize the Dynamic Pattern of Human Motion Based on Plantar Pressure Measurements

Foot plantar pressure measurements provide a window into the analysis of gait and posture and offer valuable insights into movement quality that are used in the health, sports and commodities (e.g. shoes) sectors. In this paper, we propose a novel data-driven methodology to profile subjects performing a running exercise and to identify groups with unique gait characteristics. The methodology quantifies trial-base similarities using dynamic time warping and, thereby, group different trials based on their precise spatial and temporal dynamics. The characterization of each group revealed the existence of strikingly different pressure profiles. This methodology opens the possibility to develop accurate benchmarking and prediction algorithms on homogeneous groups of running profiles and enables the followup in time of the evolution of pressure patterns during movement execution.

[1]  Martyn Shorten.  Plantar pressure distribution and footwear design , 2009 .

[2]  Tommi Kärkkäinen,et al.  Comparison of Internal Clustering Validation Indices for Prototype-Based Clustering , 2017, Algorithms.

[3]  Rezaul K. Begg,et al.  Foot Plantar Pressure Measurement System: A Review , 2012, Sensors.

[4]  S. Chiba,et al.  Dynamic programming algorithm optimization for spoken word recognition , 1978 .

[5]  Xingda Qu,et al.  Estimation of Foot Plantar Center of Pressure Trajectories with Low-Cost Instrumented Insoles Using an Individual-Specific Nonlinear Model , 2018, Sensors.

[6]  Jun Wang,et al.  Generalizing DTW to the multi-dimensional case requires an adaptive approach , 2016, Data Mining and Knowledge Discovery.

[7]  Hae-Sang Park,et al.  A simple and fast algorithm for K-medoids clustering , 2009, Expert Syst. Appl..

[8]  Uri Gottlieb,et al.  Spatiotemporal Gait Parameters as Predictors of Lower-Limb Overuse Injuries in Military Training , 2016, TheScientificWorldJournal.

[9]  Ya-Bo Yan,et al.  Normal foot loading parameters and repeatability of the Footscan® platform system , 2017, Journal of Foot and Ankle Research.

[10]  Yong Feng,et al.  A human identification method based on dynamic plantar pressure distribution , 2011, 2011 IEEE International Conference on Information and Automation.

[11]  Laurent Malisoux,et al.  Plantar pressure measurements and running-related injury: A systematic review of methods and possible associations. , 2016, Gait & posture.