Development of a BCCT Quantitative 3D Evaluation System through Low-Cost Solutions

Breast cancer is one of the most mediated malignant diseases, because of its high incidence and prevalence, but principally because of its physical and psychological invasivity. Breast Cancer Conservative Treatments (BCCT) allows a local control of the disease, with a survival similar to that obtained with a mastectomy, but with a better aesthetic result: the tumor is excised together with a small healthy tissue layer. In BCCT the surgical outcome depends on several factors, many of them difficult to assess, thus leading to a significantly heterogeneous results. For this reason, it is fundamental to evaluate specific surgical procedures on the basis of their aesthetic outcome through specific quantitative tools. The Breast Cancer Conservative Treatment.cosmetic results (BCCT.core), is a software recently developed with the objective to overcome the limitations of reproducibility and objectivity of the methods currently used to evaluate the aesthetic result of BCCT. This software is based on the comparison between the treated and non-treated breast in frontal photographs from the patients. Several indices related to the surgical aesthetic result are automatically obtained from the image, making the evaluation fast, easy and reproducible. Although the BCCT.core system presents satisfactory results, presents a significant limitation. The female breast is a complex three dimensional (3D) object and its boundaries are rather fuzzily defined in two dimensional (2D) pictures, thus making difficult the body landmarks identification. On the contrary, the use of a 3D model would allow the comparison between real geometrical characteristics of the breasts including the possibility of estimating volume and 3D surface differences, in order to plan future surgical interventions. The goal of this work is the development of a simple 3D model of a female torso, using low-cost solutions, namely: a reconstruction algorithm from two uncalibrated views, through epipolar geometry approach and making use of a Kinect sensor device. The created model will be used in an updated version of a BCCT.core to obtain a full 3D aesthetic assessment of the surgical outcome. With the inclusion of measurements extracted from the 3D model, aiming to improve the global assessment result, without increasing its complexity, as the pictures are acquired with a single camera without requiring any calibration procedure.