MR-guided conformal microwave imaging for enhanced inclusion detection within irregularly shaped volumes

Abstract

Approximately 1 in 8 women will develop breast cancer in their lifetime. Estimates suggest 230,500 new cases of invasive breast cancer in 2011, resulting in approximately 40,000 deaths. Traditional screening technologies, such as X-ray mammography use ionizing radiation and suffer from high false-positive and false-negative rates. Due to the high contrast that exists between the dielectric properties of normal and abnormal breast tissue, microwave-imaging spectroscopy has proven an attractive breast cancer imaging modality. We have shown that the incorporation of a volume’s internal structural information into our image reconstruction algorithm can increase the accuracy of recovered dielectric properties. Additionally, image reconstruction has benefited from the use of a custom reconstruction mesh generated from the imaged volume’s perimeter boundary. This information is used in a conformal microwave image (CMI) reconstruction process, and has increased the accuracy of recovered high contrast regions within the volume’s perimeter without the use of prior internal spatial information. In simulation and phantom experiments with regular geometries, boundary information is obtained through spatial measurements. For irregularly shaped boundaries, alternative means are necessary for accurate boundary extraction. In this paper we demonstrate the MR-guided CMI reconstruction process for an irregularly shaped boundary; boundary information extracted from MR images will be used to generate a custom boundary-derived mesh for microwave image reconstruction. Results from images reconstructed using the MR-guided CMI reconstruction process will be compared with uniformly reconstructed images, highlighting the increased accuracy of high contrast features within the volume without the use of prior internal spatial information.