Principal component analysis : three-dimensional asymmetry in joint kinematics of unilateral transfemoral amputees

• Main Author: Tan, Queenie Lin Ling
• Other Authors: Kong Pui Wah
• Format: Final Year Project
• Language: English
• Published: 2016
• Subjects: DRNTU::Engineering::Computer science and engineering::Computer applications::Life and medical sciences
• Online Access: http://hdl.handle.net/10356/66574
• Institution: Nanyang Technological University

Background: To cope with the partial limb loss, transfemoral amputees (TFA) adapt compensatory strategies during walking which result in imbalance loading of the prosthetic and anatomical limb and the consequential asymmetrical gait pattern. Purpose: The present study focused on evaluating key asymmetrical characteristics in three-dimensional joint kinematics of TFA compared to able-bodies using principal component analysis (PCA) so as to understand the prosthetic adaptation of these individuals. Methods: 7 unilateral TFA were invited to walk across a 10m path at self-selected speed in their own non-computerized prostheses and shoes. Participants were selected based on having at least two years amputation and ability to walk and exercise on regular basis. They were referenced to 77 healthy able-bodies for comparison. Joint kinematic data were obtained from a motion capture system and asymmetry was assessed as absolute bilateral limb difference before analysing with PCA. Results: Key characteristics of PCA revealed that only first principal component (PC1) had significant group differences and high relevance to angular kinematics. The feature of PC1 comprised of magnitude of motion throughout gait cycle and accounted for 37.2% of the variance. PC1 also revealed significant asymmetrical differences between TFA (2.90±0.97) and able-bodies (-0.21±0.41) in the three planes. Conclusions: Aside from the commonly described sagittal asymmetry, PC1 was also able to highlight significant group relevance of frontal and transverse planes to angular kinematics, specifically in the ankle transverse plane. These findings suggested that PCA can be a powerful tool in in quantifying alterations in gait pattern among TFA and able-bodies.