Vibroarthrographic Signal Spectral Features in 5-Class Knee Joint Classification

Area of Science:

• Biomedical Engineering
• Medical Diagnostics
• Signal Processing

Background:

• Vibroarthrography (VAG) offers a non-invasive approach for joint condition assessment.
• Accurate diagnosis of knee joint pathologies like chondromalacia patellae and osteoarthritis is crucial.

Purpose of the Study:

• To develop and evaluate novel spectral features for Vibroarthrography (VAG) signal analysis.
• To improve the diagnostic accuracy of knee joint conditions using VAG.

Main Methods:

• Proposed ten new spectral features and Frequency Range Maps for VAG signal analysis.
• Classified VAG signals into five groups: three chondromalacia patellae stages, osteoarthritis, and healthy controls.
• Compared novel features against state-of-the-art methods using Bhattacharyya coefficient and ten classification algorithms.

Main Results:

• The proposed spectral features demonstrated superiority over existing methods.
• Achieved over 25% improvement in Bhattacharyya coefficient and an average of 9% increase in classification accuracy.
• Effectively distinguished between all five condition classes, including neighboring stages.

Conclusions:

• The novel spectral features significantly enhance the diagnostic capability of Vibroarthrography for knee joint conditions.
• VAG, augmented with these new features, presents a promising tool for clinical joint assessment.
• Frequency Range Maps offer a valuable visualization for feature interpretation.