Low Cost 3D Scanners Along the Design of Lower Limb Prosthesis Claudio COMOTTI 1 , Giorgio COLOMBO 2 , Daniele REGAZZONI* 1 , Caterina RIZZI 1 , Andrea VITALI 1 1 University of Bergamo, Dalmine (BG), Italy; 2 Politecnico of Milano, Milan, Italy DOI: 10.15221/15.147 http://dx.doi.org/10.15221/15.147 Abstract The evolution of 3D scanning systems has determined a large range of commercial solutions available on the market with different costs depending on their performances. The most interesting scanners for the sake of this research rely on structured light optical sensors like Microsoft Kinect v1 sensor, which are extremely low-cost, but they still provide a precision that is valuable for some medical applications, e.g., the scanning of a residual limb. In this research work, we present our CAD system based on a knowledge-guided approach to design the socket starting from two different 3D acquisition systems; the first one uses Microsoft Kinect and the second one exploits MRI volumes to get the final 3D shape of residual limb. Then, a comparison of introduced techniques to create 3D shape is exposed. Final outcomes are shown and discussed in the paper. Keywords: 3D scanning, socket, MRI, Kinect, residual limb model, lower limb prosthesis 1. Introduction The growing diffusion of 3D scanning technologies represents an interesting opportunity in several application fields. We pay attention to orthopedic products that are strictly driven by human body shapes, such as for a lower limb prosthesis. Prosthesis design is composed by different steps among which the design of the socket is crucial. The socket is the most customized part because its shape must be defined according to the shape of the residual limb as well as other parameters relative to the evaluation of residual and anthropometric patient’s data. In the last decade, our research group has developed a software platform, named Socket Modelling Assistant (SMA), in order to emulate traditional manufacturing process for lower limb prosthesis. SMA makes available a set of virtual tools, which emulates operations made by orthopedic technicians during traditional workflow in an orthopedic laboratory. SMA starts prosthesis design from 3D model of residual limb, which can be obtained using either 3D portable scanner or MRI images. The whole project follows the low cost philosophy and thus, the choice of 3D scanning systems has been centered on the use of cheap 3D systems on which we can develop our custom software modules for SMA. There are different solutions that have been tested into the system, such as Microsoft Kinect v1. This type of devices can use available a wide set of low-cost or free software applications that permit to reconstruct 3D models starting from the acquired point cloud of an object. Our aim is to reach the best quality of the 3D virtual residual limb according to precision of measurement and the structure of mesh through the use of low cost IT systems and custom software modules developed within this research activity. In this paper, firstly, we introduce the differences between traditional process manufacturing and our system SMA to design the socket of lower limb prosthesis. Then, a presentation of available 3D scanner will be introduced and discussed. The use of MRI volume and 3D scanner acquisition is discussed exposing a case study that takes in consideration a transfemoral (AK) patient. Finally, preliminary test and reached results will be discussed. 2. Prosthesis development process 2.1. Traditional socket development process The prosthetic socket is the most critical component for both transfemoral and transtibial prostheses. It is manually made from a positive chalk mold of the residual limb. The plaster mold traditionally is realized using a fully handmade procedure. In this process, the prosthetist optimizes the socket shape (Figure 1) to reach a full comfortable contact with the prosthesis by the patient. The first phase consists in making some measurements of the residual limb circumferences and then it is possible to proceed with the manufacturing. In this step the technician creates a negative cast manipulating by hands plaster patches directly on patient’s residual limb. After that, s/he makes manually the positive model. This plaster model is modified by adding and removing material in particular zones, according to residual limb measurements and patient’s characteristics (e.g., muscle and soft areas). 6th International Conference on 3D Body Scanning Technologies, Lugano, Switzerland, 27-28 October 2015 147