Automatic Below-Knee Prosthesis Socket Design: A Preliminary Approach
In this work we present a preliminary study on a system able to design automatically sockets for lower-limb prosthesis. The socket is the most important part of the whole prosthesis and requires a custom design specific for the patient’s characteristics and her/his residuum morphology. The system takes in input the weight and the lifestyle of the patient, the tonicity level and the geometry file of the residuum, and creates a new model applying the correct geometric deformations needed to create a functional socket. In fact, in order to provide the right fit and prevent pain, we need to create on the socket load and off-load zones in correspondence of the critical anatomical areas. To identify the position of such critical areas, several neural networks have been trained using a dataset generated from real residuum models.
KeywordsLower limb prosthesis Neural network Prosthetic socket CAD
- 4.Haykin, S.: Neural Networks and Learning Machines – Pearson. Prentice Hall, USA (2008)Google Scholar
- 6.Jadeberg, M., Simonyan, K., Vedaldi, A., Zisserman, A.: Synthetic data and artificial neural networks for natural scene text recognition. In: Workshop on Deep Learning and Representation Learning (2014)Google Scholar