Signal calibration for an electrical impedance mammography system

Electrical Impedance Tomography (EIT) technology has been applied clinically since the 1980s. Numerous papers have addressed a variety of systematic error sources and indicated different calibration methods. The Sussex Mk4 Electrical Impedance Mammography (EIM) system has been developed for the investigation of early stage breast lesions. Investigations have shown that the system performance is subjected to a number of systematic errors: frequencies-dependant noise level due to both internal and external sources; stray capacitance within both PCB tracks and cable connections; and artefacts generated by patient movement during scanning etc. This paper reports upon several traditional and novel calibration methods utilized to reduce some of these errors in the acquired signals before image reconstruction. Techniques used include frequency spectrum analysis, filtering, phase calibration and other means of noise reduction. Results of both before and after calibration are presented and analyzed. The conclusion is reached that the signal quality of the Sussex Mk4 EIM system is such that the system is, post-calibrated, capable of producing images for the diagnosis of breast cancer.