Redundancy Reduction for Sensor Deployment in Prosthetic Socket: A Case Study.

Area of Science:

• Biomedical Engineering
• Rehabilitation Engineering
• Wearable Technology

Background:

• Amputee discomfort is often caused by irregular pressure from prosthetic sockets.
• Wearable sensors inside sockets can map pressure distribution for design improvements.

Purpose of the Study:

• To present a clustering-based analysis method for evaluating and optimizing prosthetic socket sensor density and layout.
• To reduce local redundancy in sensor deployment within prosthetic sockets.

Main Methods:

• Utilized Self-Organizing Map (SOM) and K-means for sensor data clustering.
• Employed Pearson correlation coefficient (PCC) to identify and remove redundant sensors.
• Applied Jenson-Shannon Divergence (JSD) and mean pressure for validation.

Main Results:

• A clustering-based method successfully identified sensor layout redundancy.
• Redundant sensors in posterior and medial regions were identified for reduction.
• Key pressure features were preserved after sensor optimization.

Conclusions:

• The proposed method effectively optimizes sensor configurations for intra-socket pressure measurements.
• This approach aids in improving prosthetic socket fitting and reducing amputee discomfort.