Integrating statistical shape modelling into a silhouette-based 3D reconstruction process to model a transradial defect
Steven Goes (TU Delft - Mechanical Engineering)
Dick Plettenburg – Mentor
Juan Cuellar Lopez – Mentor
Gerwin Smit – Mentor
Toon Huysmans – Mentor
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
Background: In low- and middle-income countries there is a high demand for prosthetic devices. An automatic system, in which hand prostheses are manufactured with 3D printers can potentially offer a solution for patients having a transradial defect in these areas. As part of such a system, a detailed 3D model of the residual limbs is needed. In order to make the process of creating this 3D model more accessible to low- and middle-income countries, a new silhouette-based 3D reconstruction process is observed which can be implemented in a smartphone application.
Objective:Measure the accuracy of the observed method.
Methods: A database of artificial residual limbs and an experimental algorithm is created. This algorithm consists of two parts. The first part simulates the process of capturing pictures of a residual limb with a smartphone camera. The second part performs an automatic silhouette-based 3D reconstruction.
Results: When the reconstruction method is performed on a known residual limb shape with three silhouette images, the highest measured 3D reconstruction accuracy is 1.12 ± 0.57 mm. When the method is performed on an unknown residual limb shape with three silhouette images, the highest accuracy is 6.48 ± 2.15 mm.
Conclusion:This work presents a technique for reconstructing a residual limb by means of silhouette images. The observed method can be considered as a promising 3D reconstruction approach for prosthetic designing. The method could be improved by having access to a larger database of residual limb shapes and by analysing and finding the optimal input arguments for the optimiser.
Clinical Relevance:The observed method provides a low-cost and accessible approach to model a residual limb for the design of a fitting prosthetic socket that can be manufactured by a 3D printer.