Metrics for Assessing the Similarity of Microwave Breast Imaging Scans of Healthy Volunteers

Abstract

Microwave radar imaging is promising as a complementary medical imaging modality. However, the unique nature of the images means interpretation can be difficult. As a result, it is important to understand the sources of image differences, and how much variability is inherent in the imaging system itself. To address this issue, we compare the effectiveness of six different measures of image similarity for quantifying the similarity (or difference) between two microwave radar images. The structural similarity index has become the de facto standard for image comparison, but we propose that useful information can be acquired from a measure known as the Modified Hausdorff Distance. We apply each measure to image pairs from sequential scans of both phantoms and volunteers. We find that rather than using a single value to quantify the image similarity, by computing a number of values that are designed to capture different image aspects, we can better assess the ways in which the images differ.

Adaptive Monostatic System for Measuring Microwave Reflections from the Breast

“…For each antenna position, we introduce randomly distributed perturbations with mean of 0 and standard deviation of 2 units to each of the 4 degrees of freedom. Even with this significant perturbation, the images were virtually visually indistinguishable, had no displacement of the main response centroid, and the difference in the overall backscatter response measured using the Modified Hausdorff Distance (MHD) was 0.28 mm [ 17 ]. With systematic shifts in antenna positions, similar consistency in images was also noted.…”

“…Additionally, the difference in the overall backscatter response measured using the MHD as described in B.R. Lavoie et al [ 17 ] ranged between 0.49 mm and 0.17 mm. The highest and lowest values were between scans 1–3 and 2–3, respectively.…”

Assessing Patient-Specific Microwave Breast Imaging in Clinical Case Studies

“…In addition to patient trials, both analytical and experimental investigations of the fundamentals of microwave breast imaging have also continued, examining factors such as antenna layout, artefact removal, imaging algorithm selection, prior information integration, multi-modality imaging, contrast agents and phantom development [ 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 ]. In the past two years, open-source imaging tools such as the MERIT toolbox, introductory textbooks such as that published by Nikolova and now open-source experimental data have been made available to the microwave imaging community [ 1 , 29 , 30 ].…”

“…Many of the challenges faced when imaging patients identified from the clinical studies to date can be categorised in four broad areas: Inefficient coupling of energy into the breast [ 31 ]; Imaging domain changes during acquisition [ 31 , 32 , 33 ]; Intrapatient variation due to the menstrual cycle, hormonal changes or weight differences [ 33 ]; Interpatient variation in breast size, shape and composition [ 10 , 18 , 21 ]. …”

Optimal microwave breast imaging using quality metrics and simulated annealing algorithm

“…While tomography aims at reconstructing the dielectric distribution within the breast, [12][13][14] radar-based imaging focuses on locating the strong scattering area. [15][16][17] In radar-based imaging, confocal microwave imaging (CMI) is one extensive used method to transform the microwave reflections signals into reconstructed images of the breast. 18 In CMI method, the image is reconstructed by estimating the propagation delay of the signal and adding signals from different channels together.…”