Lifespan Prediction Using Deep Learning for a Transtibial Socket Made by Natural Fiber Reinforced Composite Material

The primary goal of this study is to leverage artificial intelligence, specifically deep learning, to predict the lifespan of trans-tibial prosthetic sockets made from natural fiber (alfa). Prosthetic sockets are vital components that interface directly with the residual limb, and their durability is crucial for the comfort and functionality of the prosthesis. Traditional methods of predicting socket lifespan often rely on empirical data and manual assessment, which can be both time consuming and imprecise. To address this, our study integrates AI with experimental data, employing a deep learning model that analyzes angular deflection cycles, which represent cyclic stress patterns. This model incorporates design and material parameters, and is trained on a comprehensive dataset including deflection cycles, material responses, and failure data. By correlating angular deflection behavior with long term performance, our approach significantly enhances the accuracy of durability predictions. This method not only improves patient outcomes and prosthetic performance but also facilitates better design and manufacturing strategies, leading to more reliable and enduring prosthetic devices.