Weighted delay-and-sum beamformer for breast cancer detection using microwave imaging

Many similarities exist between ultrasound and microwave imaging, however, ultrasound is typically performed with a handheld transducer whereas antennas are placed around the subject of imaging for microwave breast imaging. Hence, ultrasound beamformers are not always directly applicable for microwave imaging. Signals traverse different paths experiencing different delays, attenuations, and even different expected shapes of signal due to dispersion. Besides, each signal may contain the response from a different point of the boundary of the object. In this work, increased coherencies between the signals of certain antenna pairs are shown statistically and these characteristics exploited to develop a novel beamformer which weights the received signals prior to imaging using these characteristics. Specifically, a novel data-independent beamformer is proposed by weighting signals based on the relative positions of the antennas. These weights are relevant to the intrinsic worth of signals according to the information of the tumor response. This patient-independent weighting can make the delay-and-sum (DAS) beamformer more robust. The contrast is improved as well, with an absolutely negligible computational load compared to DAS. Weighted-DAS (WDAS) is evaluated using realistic breast phantoms and the improvement of the output quality in comparison to DAS is shown.

Automated summary

Ultrasound and microwave imaging share similarities, but due to differences in signal paths, delays, and attenuations, ultrasound beamformers are not directly applicable to microwave imaging. This study demonstrated increased coherencies between certain antenna pairs and developed a novel beamformer that weights the received signals based on these characteristics to improve imaging quality.