Using Electrical Impedance Tomography and Machine Learning for Breast Cancer Detection

Breast cancer is a significant health concern for women worldwide. There are several techniques and screening methods available for the detection of breast cancer including mammography, ultrasound imaging, and magnetic resonance imaging (MRI). Each of these techniques plays a crucial role in identifying cancerous cells or tissue abnormalities, but they also have limitations and potential drawbacks. On the other hand, electrical impedance tomography (EIT) is a radiation-free, cost-effective, and non-invasive imaging technique. It utilizes low frequency signal to measure the electrical impedance of the internal body tissues. By applying this technique, it is possible to reconstruct a tomographic image that provides valuable insights into the internal composition, properties, and abnormalities of the object under test. In this paper, we aim to investigate the potential use of EIT, alongside machine learning and image reconstruction techniques, for breast cancer diagnosis. The integration of EIT with machine learning techniques opens new possibilities for early detection, improved accuracy, and personalized treatment planning. The contribution of this work is to provide to provide a portable and effective device that can potentially be used for early breast cancer detection. The outcome of this study is a portable and automated electrical impedance tomography device with a machine learning model trained on data obtained from chicken phantom, mimicking breast tissue. The obtained accuracy reached 78.75%.