Simulation of 2 Dimensional Non-invasive Electrical Impedance Tomography System for Early Stage Breast Cancer Detection

A method to determine the impedance distribution in any area of the body is electrical impedance tomography (EIT). Based on the idea that different tissues have different electrical characteristics, EIT is employed in medical imaging. Tissues has specific resistivity value. The resistivity of tissue changes with blood perfusion. The cancer tumor is more vascularized i.e. it receives more blood supply than the surrounding normal tissue, hence resistivity of such tissues is less. EIT technique's foundation is the use of the differential in resistivity between healthy and cancerous breast tissue as a cancer indicator. By applying an alternating current to contact electrodes arranged in different patterns on an object's surface, boundary potentials are measured and the spatial distribution of electrical conductivity inside the object (particularly the human body) is imaged using EIT. It is non-invasive, relatively cheap, and safe medical technique. Image Reconstruction in EIT involves solution to inverse problem. This research involves simulation of forward problem in COMSOL Multiphysics software to create the experimental setup. Further image reconstruction is carried out using different reconstruction techniques as an inverse problem. Various reconstruction algorithms like Gauss Newton, Tikhonov Regularization, NOSER Algorithm, Total Variation Reconstruction (TV) and Laplace algorithm, are compared. Finally Monte Carlo method is used to locate and find the size of the impurity (signify the tumor) added. From the above mentioned algorithms NOSER Algorithm gives better reconstruction results. The percentage of error for NOSER algorithm is 0.66% for conducting impurity of 0.8 cm.