Principal Component Analysis (PCA) is a useful method for analysing human motion data. The objective of this study was to use PCA to quantify the biggest variance in knee kinematics waveforms between a Non-Pathological (NP) group and individuals awaiting High Tibial Osteotomy (HTO) surgery. Thirty knees (29 participants) who were scheduled for HTO surgery were included in this study. Twenty-eight NP volunteers were recruited into the study. Human motion analysis was performed during level gait using a modified Cleveland marker set. Subjects walked at their self-selected speed for a minimum of 6 successful trials. Knee kinematics were calculated within Visual3D (C-Motion). The first three Principal Components (PCs) of each input variable were selected. Single-component reconstruction was performed alongside representative extremes of each PC to aid interpretation of the biomechanical feature reconstructed by each component.Abstract
Objectives
Methods
Exploring the relationship of gait function pre and post total knee replacement (TKR) in two groups of patients. Three-dimensional gait analysis was performed at Cardiff University, UK, and Karolinska University Hospital, Sweden, on 29 and 25 non-pathological (NP) volunteers, and 39 and 28 patients with end-stage knee osteoarthritis (OA), respectively. Patients were assessed pre and one-year post-TKR. Data reduction was performed via Principal Component (PC) analysis on twenty-four kinematic and kinetic waveforms in both NP and pre/post-TKR. Cardiff's and Karolinska's cohorts were analysed separately. The Cardiff Classifier, a classification system based on the Dempster-Shafer theory, was trained with the first 3 PCs of each variable for each cohort. The Classifier classifies each participant by assigning them a belief in NP, belief in OA (BOA) and belief in uncertainty, based on their biomechanical features. The correlation between patient's BOA values (range: 0–1, 0 indicates null BOA and 1 high BOA) pre and post-TKR was tested through Spearman's correlation coefficient in each cohort. The related-samples Wilcoxon signed-rank test (α=0.05) determined the significant changes in BOA in each cohort of patients. The Mann-Whitney U test (α=0.05) was run to explore differences between the patients’ cohorts.Abstract
Objectives
Methods
Exploring the association of objective lower limb function pre and post total knee replacement (TKR). 3D gait analysis was performed on 28 non-pathological participants (NP) and 40 patients with advanced knee osteoarthritis (OA) before and approximately one year after TKR. For NP and OA patients pre/post-TKR, 12 waveforms on kinetic and kinematic variables of the operative side were chosen to perform data reduction through Principal Component (PC) Analysis. The Cardiff Classifier, a classification system based on Dempster-Shafer theory, was trained with the first 3 PCs of each variable. The 18 highest-ranking PCs classifying the biomechanical features of each participant as Belief in Healthy, Belief in OA (BOA) or Belief in Uncertainty were used to quantify biomechanical changes pre- to post-TKR. The correlation between patients’ BOA values (range: 0 to 1, 0 indicates null BOA and 1 high BOA) pre- and post-TKR was tested through Spearman's correlation coefficient. Wilcoxon matched-pair test (α<0.05) determined the significance of the change in BOA.Abstract
Objectives
Methods
