AI Algorithms bring new design

TWI Hellas (a fully owned subsidiary of TWI Ltd, UK), in the context of SocketSense project, develops state-of-the-art algorithms based on AI tools, performs model-driven data analysis and builds digital simulation models, in order to design a socket having improved fitness and greater comfort. TWI-Hellas, also has the role of the technical manager of SocketSense project and is responsible for the integration of all hardware and software systems. In the following paragraph, we present a more detailed description of the work in progress on AI tools for improving prosthetic socket comfort.

We are currently developing a software that generates an improved 3D digital model of the amputee’s socket, offering better socket-fitness and greater comfort to the amputee. The digital process is based on a fuzzy-logic inference engine, which accepts as input the point-cloud of the 3D digital model of the socket (see figure 1) and the real-time pressure data generated by the sensors embedded in the actual check-socket that is worn by the user (in static and dynamic conditions). A set of fuzzy rules encapsulating the expert’s knowledge (the prosthetist) generates an output that comprises a set of rectification actions i.e. specific socket shape modifications in order to redistribute the pressures on the stump and lead to a better fitting of the socket. The rectification actions are applied using transformation functions and result in a new rectified 3D digital model of the socket (see figure 1).

Figure 1
Fuzzy-logic inference engine for improving the fitness and comfort of a prosthetic socket.

Our goal is to use the software as a decision support tool for the prosthetist on making decisions on the final shape of the socket. The iterations needed for progressing from the check-socket to the final socket will be reduced and the rectifications are expected to be superior to the state-of the-art since, the fuzzy algorithm not only does it take into account the experts’ knowledge but also processes the pressure data generated in the socket during donning and gaiting.


TWI Hellas is based in Greece

They are a membership-based organization, supporting both individuals and companies alike. We exist to provide authoritative and impartial expert advice, know-how and safety assurance through engineering, materials and joining technologies – helping you design, create and operate the best products possible.