Spectral capacitively coupled electrical resistivity tomography for breast cancer detection

Early detection of breast cancer is key for a better survival rate, so there are huge demands for a safe and reliable breast cancer screening method, which currently is not available. Earlier studies are showing that the impedance spectroscopy can be an effective discriminator for breast tumor against healthy tissues. Traditional electrical impedance tomography (EIT) has progressed well into clinical stages for such an application with some promising results. But EIT generally operates at low frequencies, ranging typically lower than 1 MHz. Previous research work has suggested that the important information on the tumor impedance spectra exists in the higher frequency domain. Thus, this work focuses on the introduction and the evaluation of a spectral capacitively-coupled electrical resistivity tomography (CCERT) for breast cancer diagnosis. Compared to the traditional EIT, CCERT can reach a much higher frequency domain producing the spectroscopic images in a broader range. This in turns offers improved discrimination of tumors against normal tissues via frequency difference imaging or spectral imaging as it is shown in this paper. Additionally, direct contact between the EIT electrodes and the surface of the body, which causes significant challenges and uncertainties in measured data, is avoided in CCERT inherently a contactless method. In this paper, a 3D in-silico model of the CCERT system is created and solved using the finite element method (FEM) to simulate the forward model for each frequency. The feasibility simulation analysis is conducted based on the existing clinical breast cancer data and a CCERT reconstruction. For the experimental verification, an 8-electrode laboratory phantom was tested with the excitation frequency ranging from 200 kHz to 13 MHz. The simulation and experimental results show the promising potential of this method for breast cancer imaging. The wide-range spectroscopic images provide better discrimination of tumors against healthy tissues through their spectral conductivity profile as well as the rate of changes in spectral conductivity images.