Matching Pursuit for Inter-Scatterer Spacing Estimation from Ultrasound RF Echo Signals

Xiuhua Zeng,Yufeng Zhang
DOI: https://doi.org/10.1145/3301879.3301895
2018-01-01
Abstract:The pathological changes in the diseased tissues could alter the spatial organization of the microstructures of their normal tissues. Ultrasound RF echo signals carry the microstructure information of the tissues whose regular scatterer spacing (SS) can be analyzed and applied as a signature for the quantitative diagnosis of diseases. In this paper, we propose a novel approach based on the matching pursuit (MP) algorithm to improve the ISS estimation from the ultrasound RF echo signals. The signals are decomposed into a linear combination of the Gabor atoms, and then the atoms highly matched the regular component are selected based on the critical point in the curve of the first order difference (FOD) of the reconstructed signal to residue ratio (RSRR). Accordingly, the ISSs can be estimated with the adjacent location indexes of the selected Gabor atoms. Results of simulation experiments with the MP-based and WT-based methods are presented. The results demonstrate that the MP-based method can improve the ISS estimation, especially for those simulated signals with low regular components, and potentially provide more accurate information for the disease diagnosis.
What problem does this paper attempt to address?