Design and Development of An Application For The Pre-Operative Planning of DIEP-Based Breast Reconstruction

Breast cancer is the most common type of cancer among women (excluding skin cancer) and corresponds to the second deadliest cancer type in this gender. When women are diagnosed with breast cancer, they have a plethora of options, but, in some cases, the only solution is a mastectomy. This procedure consists of completely removing the breast tissue, and sometimes both breasts must be removed, which can lead to physical insecurities and even psychological distress. Breast reconstruction surgery may help women overcome these problems. Among all the possible reconstruction procedures, a common solution is the use of autologous tissue flaps, which give a more natural look and feeling and last longer when compared to implants. Tissue can be collected from different body regions, but the current gold standard is the use of the Deep Inferior Epigastric Perforator (DIEP) flaps, which provide proper blood supply and decrease abdominal wall morbidity. Extensive analysis of images, obtained by Computerized Tomographic Angiography (CTA), is required to retrieve the abdominal perforators' characteristics, such as coordinates relative to the navel, caliber, and orientation. This helps to locate the tissue with an optimal blood vessel network for the procedure. This is a challenging and time-consuming process, nevertheless it is essential, since it has a strong impact on the success rate of the procedure. It is currently done by a technician or a radiologist who documents the relevant information about the perforators in a report, which is then provided to the surgeon who will perform the procedure. To alleviate and even improve the task of detecting the blood vessels, a semi-automatic Computer Aided Detection (CAD) algorithm has been developed. The algorithm divides the detection problem into two parts: the detection of the subcutaneous portion of the perforators and the detection of the intramuscular portion of the perforators. Both of these parts require user input, hence being semi-automatic. In the subcutaneous detection, the user must make two inputs at the end of each perforator to get its direction, and in the intramuscular part, the user must input the location where the Deep Inferior Epigastric Artery (DIEA) perforates the posterior sheath of the muscle. In the end, a report containing information about the perforators is generated. Even though this algorithm is fully developed and working, medical professionals cannot utilize it since it lacks a user interface. The goal of this thesis is the design and development of a desktop application that can be used fully and independently by the medical staff. To achieve this goal, the algorithm must be translated from its original language, MatLab, to Python, and a requirements analysis must be conducted to determine functionalities, as described by clinicians, that the application should include. Finally, the application developed should have an easy-to-understand and user-friendly interface, allow to see the 3D path of the perforators and, have a modular design that allows adding new functionalities