Quantifying the Effects of Increasing Mechanical Stress on Knee Acoustical Emissions Using Unsupervised Graph Mining

Area of Science:

• Biomechanics
• Acoustics
• Data Mining

Background:

• Mechanical stress on knee joints can lead to injury.
• Assessing knee joint loading is crucial for rehabilitation and injury prevention.
• Non-invasive methods for monitoring joint loading are needed.

Purpose of the Study:

• To investigate the relationship between mechanical stress and knee acoustical emissions.
• To analyze knee sounds using unsupervised graph mining.
• To develop a method for quantifying joint loading based on acoustic signals.

Main Methods:

• Miniature contact microphones were placed on four knee locations (patella, meniscus).
• Audio features were extracted from acoustical signals.
• A graph community factor (GCF) was calculated using k-nearest neighbor graphs and Infomap community detection.

Main Results:

• The GCF significantly increased with vertical loading forces in 12 healthy subjects.
• Increased sound complexity correlated with increased joint forces.
• Medial patella and lateral meniscus microphone placements showed the most sound variation.

Conclusions:

• Knee acoustical emissions complexity reflects joint loading.
• The GCF can quantify knee joint loading.
• Optimal microphone placement can enhance sensitivity to loading for future applications in rehabilitation and activity monitoring.