Study on a Novel Localization Approach of Breast Tumor based on Electrical Impedance Tomography

A novel approach of breast tumor localization based on electrical impedance tomography (EIT) is proposed. Under the Comsol+Matlab simulation platform, a circular truncated cone model with two layers of electrode array is constructed, the top layer is configured with 8 electrodes and the bottom layer with 12. The neighboring boundary voltages are measured when a pair of neighboring exciting electrodes inject a low-amplitude alternative current signal. An improved Tikhonov regularization algorithm is adopted to reconstruct the image. According to the imaging results of each layer, the height of the tumor can be estimated, and K-means clustering algorithm is utilized to achieve the central position of the tumor, thus the location of the tumor can be acquired. The simulation study indicates that the tumor can be identified with errors of height under 5mm and center between 0.5mm to 2mm.