Static and dynamic pressure prediction for prosthetic socket fitting assessment utilising an inverse problem approach

Area of Science:

• Biomedical Engineering
• Artificial Intelligence in Medicine
• Prosthetics and Orthotics

Background:

• Current lower-limb prosthetic socket fitting requires improved tools.
• Accurate measurement of limb/socket interface pressures is crucial for optimal prosthetic function and patient comfort.

Purpose of the Study:

• To design and clinically test an artificial intelligence (AI) approach, specifically inverse problem analysis, for determining pressures at the limb/prosthetic socket interface.
• To develop a tool that aids in achieving a "right first time" socket fit.

Main Methods:

• Developed a backpropagation artificial neural network (ANN) to predict interfacial pressures from strain data.
• Validated the ANN using strain data collected from the socket surface of a unilateral trans-tibial amputee.
• Employed inverse problem analysis to calculate boundary conditions generating known strain.

Main Results:

• The ANN demonstrated an 8.7% difference when comparing predicted interfacial pressures to actual pressures.
• The validated ANN accurately mapped static and dynamic interfacial pressure distribution during ambulation and varying axial loads.
• The AI methodology provides a full-field study of pressure distribution within the prosthetic socket.

Conclusions:

• A novel methodology for quantitative analysis of prosthetic socket pressures in a clinical setting has been established.
• This AI-driven approach enhances socket fitting precision, leading to improved patient comfort.
• The method reduces fitting time and costs for prosthetists, optimizing the prosthetic care process.