Examining cranial cruciate ligament deficiency and its effect on canine gait using Parafac2

Gait analyses are typically concerned with comparing subjects’ waveform patterns for clinical or experimental reasons; two-mode data matrices (e.g., subjects × gait percents) are commonly analyzed by principal component analysis (PCA). Typical gait data, however, are measured in more than two modes (e.g., subjects × gait percents × joints). If PCA is used for such multi-mode gait data, the data need to be reorganized into a two-mode matrix, by either taking average over one mode or picking one condition in one mode. The resulting PCA solution will represent only the mean structure or a particular aspect relevant to the chosen condition. Parafac and Parafac2 have been successfully applied to three- and four-mode gait waveform data. Whereas previous Parafac2 analyses used experimental gait data, this study presents a successful application to cranial cruciate ligament (CCL) deficient and healthy canine gait data that include incomparable waveforms. Our study focuses on (a) investigating differences and interconnections between the contralateral and ipsilateral limbs of the CCL-deficient dogs; (b) determining the interactions between joints and how they vary over the gait cycle and by waveforms; and (c) providing clinical diagnostic tools using the Parafac2 solution. An optimal Parafac2 solution was able to systematically distinguish CCL-deficient dogs from healthy ones, uniquely characterize contralateral and ipsilateral limbs of the CCL-deficient dogs, and understand how the CCL deficiency affects specific joints and influences specific waveforms during specific phases of the gait cycle.

Name thesis:degree_name

M.A.

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Psychology

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2012

Author dc:creator

• Bokhari, Ehsan

Contributors dc:contributor

• Hong, Sungjin

• Cranial Cruciate Ligament (CCL) deficiency
• Canine Gait Analysis
• Parallel Factor Analysis
• Multiway Component Analysis
• Parafac2

Statement dc:rights

• Copyright 2012 Ehsan Bokhari

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/32024

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/32024