Abstract-In this work we examine, for the first time, the use of classification algorithms for earlystage tumor detection with an experimental time-domain microwave breast screening system. The experimental system contains a 16-element antenna array, and testing is done on breast phantoms that mimic breast tissue dielectric properties. We obtain experimental data from multiple breast phantoms with two possible tumor locations. In this work, we investigate a method for detecting the tumors within the breast but without the usual complexity inherent to image-generation methods, and confirm its feasibility on experimental data. The proposed method uses machine learning techniques, namely Support Vector Machines (SVM) and Linear Discriminant Analysis (LDA), to determine whether the current breast being scanned is tumor-free. Our results show that both SVM and LDA methods have promise as algorithms supporting early breast cancer microwave screening.
Abstract-We experimentally demonstrate a low-cost hardware technique for synthesizing a specific electromagnetic pulse shape to improve a time-domain microwave breast imaging system. A synthesized broadband reflector (SBR) filter structure is used to reshape a generic impulse to create an ad-hoc pulse with a specifically chosen frequency spectrum that improves the detection and imaging capabilities of our experimental system. The tailored pulse shape benefits the system by improving the level of signal detection after transmission through the breast and thus permits higher-resolution images. We report on our ability to use this technique to detect the presence of tumours in realistic breast phantoms composed of varying quantities of glandular tissue. Additionally, we provide a set of images based on this experimental data that demonstrates the increased effectiveness of the system using the SBR-shaped pulse in the localisation and identification of the embedded tumour.
Microwave radar and microwave-induced thermoacoustics, recently proposed as promising breast cancer detection techniques, each have shortcomings that reduce detection performance. Making the assumption that the measurement noises experienced when applying these two techniques are independent, we propose a methodology to process the input signals jointly based on a hypothesis testing framework. We present two test statistics and derive their distributions to set the thresholds. The methodology is evaluated on numerically simulated signals acquired from 2-D numerical breast models using finite-difference time-domain method. Our results show that the proposed dual-modality approach can give a significant improvement in detection performance.
Abstract-In this work we examine several sources of measurement uncertainty that can hinder the use of time-domain microwave techniques for breast imaging. The effects that are investigated include those due to clock and trigger jitter, antenna movements, discrepancies in antenna fabrication, and random measurement noise. We explore the significance of the noise contribution of each effect, and present methods to mitigate them when possible and necessary. We demonstrate that, after applying the aforementioned methods, the noise is minimized to the noise floor of the system, thereby enabling successful tumor detection.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.