Searching for the Optimal Frequency to Identify Breast Tumors Using Impedance Spectroscopy

Breast cancer becomes one of the biggest public health problems all around the world. Visualization of tissue electrical properties in low frequency spectrum can be highly promising detection technique of tumor with high sensitivity at an early stage. The aim of the present study was to assess the clinical utility of impedance spectroscopy for breast tissues as additional prognostic marker of cancer. Electric field simulation in frequency domain for coaxial electrode system was performed using dielectric properties of normal and pathological tissues obtained from verified data of previous studies, based on dielectric relaxation spectroscopy. Relative permittivity and conductivity are frequency dependent parameters. To predict the dielectric behavior of breast tissues in frequency range below 1 MHz two-pole Cole-Cole relaxation model, which fits well with the limited number of frequency points, was used. There are various approaches to automatically find the coefficients of fitted curves. Particle swarm optimization, in contrast to Levenberg-Marquardt algorithm, was able to find suitable values by reducing error function to a predefined threshold. Obtained values with standard error were added to simulation model of coaxial electrode system. The finite element method was used to simulate electric field propagation in model. Frequency dependence of the impedance magnitude and phase angle for all breast tissues was plotted and analyzed. Magnitude can’t be used as criterion for tissue classification due to overlapping error bars, while phase angle values in short frequency range are clearly different for adipose, normal tissues and tumor. Hence, impedance spectroscopy can identify breast tumors.