Modeling and Simulation for Breast Conserving Therapy

Breast cancer is the most common type of cancer affecting women throughout the world and the second most common type of cancer in the United States. Significant progress have been made in the last decades in terms of early detection of breast cancer and breast cancer care, improving effectively its survival rate. As a consequence of the increased number of early detections, recent advances in breast cancer care and in particular breast conserving therapy have been made possible in order to minimize the impact of the surgery on the quality of life of patient. However, in the case of breast cancer, the outcome of the surgery, \textit{i.e.} the contour of the breast, remains potentially non-satisfactory to some patients.

The goal of the work presented in this dissertation is to provide a computational framework to model the effects of the breast conserving therapy on the breast. This computational tool aims to facilitate the dialogue between the surgeon and the patient by providing a vizualization of the breast contour following surgery and specifically to the anatomy and pathology of the patient. We developed a patient-specific, multiscale model, taking into account the anatomy of the breast of the patient, the mechanics of the breast tissues, and the biology of the wound healing following the surgery. We particularly focus in this work on an agile and modular development of this model. We provide a class of biological models of wound healing, with uses ranging from a theoretical model aiming to help understand the dynamic of the healing of breast cancer surgery wounds, to a computationally efficient model well adapted for clinical use.

In addition, we develop in this dissertation a clinical protocol designed for the validation of this multiscale model on patient data, and we present the results of a first case study with a patient undergoing breast conserving therapy. We show promising first results in the validation of some of the critical parameters of the model developed in this dissertation, and aim to define a new computational framework relevant in a clinical context.