Evaluation and development of FacePrint: A tool for the design of facial prostheses

Abstract: FacePrint is a project developed by a consortium that aims to efficiently provide facial prostheses, mainly nose and ear prostheses. One of the companies involved in the project is Purple Scout. Their task was to develop the interface that the clinicians will use to design the prosthesis. The idea is that patient and doctor sit together and decide what the prosthesis should look like using the interface. A prosthesis is generated using machine learning and then modified using parameters in the interface. When the prosthesis is designed, the data is sent to another company that 3D prints it. The entire process is estimated to take 72 hours, corresponding to a reduction of 96 \% compared to the standard way of making facial prostheses. As technology is advancing, so will FacePrint. fiXR is a continuation of FacePrint, adapted into AR (Augmented Reality). The clinician wears an AR headset and designs the prosthesis with spatial computing. When designing interfaces it is important to keep in mind that the interface needs to be usable. Usability means that the user easily and efficiently accomplishes what they aim to do with the product. This master thesis project aimed to design and improve the current interface of FacePrint. The purpose was also to explore what opportunities there are for FacePrint in AR, also known as fiXR. The project was split into three phases, where Phase 1 and 2 concern FacePrint and Phase 3 fiXR. In Phase 1 and 2 an iterative design process was used to evaluate and further develop the interface. In Phase 3 fiXR was designed and a prototype was generated. Three different test methods were used in Phase 1 and 2, and in total 25 test persons participated. There was a big improvement between the two phases and the interface became much more usable. Through different methods, a prototype made in AdobeXD was created for AR. In conclusion, the approach of splitting the project in three phases and using many different methods resulted in a usable, well liked interface with a lot of potential, as well as an intuitive AR prototype for future possibilities.