Wound model reconstruction from three-dimensional skin surface imaging using the convex hull approximation method

Ulcer wound refers to a wound with underlying medical conditions that prevent healing. The ability to measure objectively early therapeutic response is important for wound management. Early therapeutic efficacy is best assessed by measurement of wound depth and volume. This study presents a non-invasive technique for assessing ulcer wound volume from three-dimensional (3D surface scans. The accuracy of volume computation is dependent on the performance of the solid reconstruction of the wound. However, it is difficult to reconstruct solid models of wounds such as leg ulcers that extend over a large area along leg curvature and when irregularities exist on the skin surrounding the wound. The convex hull approximation preceded by surface division of the wound surface scan is proposed for solid reconstruction of wound models. This approach enables the reconstruction of models for large wounds without being affected by the surrounding irregularities. The performance of the algorithm is compared against reference volumes of wound models developed on AutoCAD. The best results are achieved using convex hull preceded by 20 surface divisions with volume computation error from 0 to 7%. Furthermore, moulds of wounds are developed to compute volumes using invasive and non-invasive methods, their percentage differences are below 8%.