Modeling Procedures for Breast Cancer Diagnosis based on Clinical Elastography Images

Nowadays, breast cancer is considered the second cause common cancer type of women death. To determine the proper therapeutic procedures before cancer spreading, early detection of cancer is a definitive step. Ultrasound elastography is considered one of the early effective noninvasive diagnostic tools. It has many advantages as low cost, its safety and the highly increasing development in various medical imaging applications.In this work, 3D modelling and simulations using virtual phantoms that were designed based on realistic in-vivo experimental results. The models were constructed for each in-vivo individual case assuring the biomechanical features of the breast tissue. The models are integrated several breast tumor's parameters including size, shape, and position. In particular, mathematical and computational analyses were used to compare this work's results by assorted specifics of in-vivo elastograms. Tumor discrimination; either malignant or benign, was performed depending on the non-linear biomechanical properties of breast tumors. To calculate the main classification parameters, tissue deformations and strain differences among the suspected mass and the normal surrounding background tissue were analyzed and empirically fitted. The results show a kindly agreement between the model outputs and the in-vivo diagnostics elastograms. Generally, the introduced finite element modeling method can be considered as a non-invasive diagnostic procedure in an early stage to preceding classify breast tumors. The 3D simulation results can assure a more theoretical insight on the behavior of nonlinear biomechanical properties that might not be obvious or convenient using clinical experimentations.