A footprint-based approach for the rational classification of foot types in young schoolchildren

Abstract

Background Discrepancy in the classification criteria among footprint parameters complicates attempts for rational classification of feet. Objective To develop a footprint-based classification technique for the rational classification of foot types by allowing simultaneous use of several parameters (co-classification). Method Static standing footprints were recorded from 132 schoolchildren. The Arch Index (AI), Martirosov's K Index (KI), Footprint Angle (FPA) and Chippaux-Smirak Index (CSI) were determined. k-means cluster analysis was applied to obtain individual classifications and co-classifications. Results Identification across classified foot types coincided with all parameters for 0.8–3.5% of the sample. Values ranged from 2.3 to 10.6% when classification coincided with three out of four parameters. The inconsistency between every initial individual classification and co-classifications corresponded to 15.2% for AI, 49.2% for KI, 41.7% for FPA and 48.5% for CSI. Conclusions The co-classification process indicated low percentages of correctly identified foot types from all parameters that suggest the dependence of the classification on the parameter of choice used to assess arch configuration. The AI gave the lowest percentages of misclassified cases during the co-classification process. The co-classification model with the 4-cluster solution is proposed and confidence limits are reported for a rational classification of feet in young schoolchildren.