Detection of Breast Cancer Using Microwave Imaging and Machine Learning-finite Element Method Inverse Models

According to Canadian Cancer Society, breast cancer is the most frequently diagnosed cancer in women. Our ability to potentially detect breast cancer in an early stage has potential of significantly decreasing mortality and hence is a very important issue in healthcare. Currently, mammography has been used widely for screening women over 50 who are statistically more vulnerable. However, it suffers from some limitations such as false negative and positive results, using ionizing radiation and patient’s discomfort. Microwave imaging has been introduced recently to overcome drawbacks of this method. However, due to the non-heterogeneous structure of the breast tissue, antenna noise and limited number of antennas that can be used (hence reducing spatial resolution) the performance of detection (and consequently estimation algorithms) can deteriorate significantly. Therefore classical inverse models based on finite-element methods are rather computationally intense due to the large number of iterations. Recently, it has been proposed that machine learning algorithms can provide sufficiently accurate results at the much lower computational times due to the fact that the training is done before the patient’s data is processed. In this paper we propose convolution neural network (CNN) configuration for inverse solutions and demonstrate its applicability using numerical results.