Segmentation and artifact removal in microwave-induced thermoacoustic imaging.

Abstract

Microwave-induced thermoacoustic (TA) imaging combines the soft-tissue dielectric contrast of microwave excitation with the resolution of ultrasound for the goal of a safe, high resolution, and possibly portable imaging technique. However, the hybrid nature of this method introduces new image-reconstruction challenges in enabling sufficient accuracy and segmentation. In this paper, we propose a segmentation technique based on the polarity characteristic of TA signals. A wavelet analysis based method is proposed to identify reflection artifacts as well. The time-frequency feature of the signal is used to assist differentiating artifacts. Ex vivo verification with experimental data is also provided.