Reduction, classification and ranking of motion analysis data: an application to osteoarthritic and normal knee...

Area of Science:

• Biomedical Engineering
• Clinical Biomechanics
• Data Science in Healthcare

Background:

• Clinical decision-making in motion analysis faces challenges with complex data interpretation and quantitative comparisons.
• Existing methods struggle with large datasets, subjectivity, and visualization needs in motion analysis.

Purpose of the Study:

• To overcome obstacles in clinical motion analysis using a novel hybrid approach.
• To characterize differences between osteoarthritic (OA) and normal (NL) knee function.
• To identify key variables for classifying knee function using motion analysis data.

Main Methods:

• A hybrid approach integrating principal component analysis (PCA), Dempster-Shafer (DS) theory of evidence, and simplex plots was developed.
• The method was applied to analyze and differentiate between osteoarthritic (OA) and normal (NL) knee function data.
• Results were compared against artificial neural network analyses for validation.

Main Results:

• The hybrid approach successfully characterized differences in knee function between osteoarthritic and normal subjects.
• A hierarchy of discriminatory variables was established for improved classification.
• The hybrid method demonstrated comparable or superior performance to artificial neural networks.

Conclusions:

• The hybrid PCA-DS approach offers a robust solution for analyzing complex motion analysis data in clinical settings.
• This method enhances the objectivity and quantitative comparison of temporal waveform data for better decision-making.
• The approach aids in identifying key biomechanical markers for osteoarthritis detection and classification.