Constrained least squares filtering for ultrasound image deconvolution

Wee Soon Yeoh; Cishen Zhang; Ming Chen

doi:10.1109/IEMBS.2005.1616801

Constrained least squares filtering for ultrasound image deconvolution

Conf Proc IEEE Eng Med Biol Soc. 2005:2005:1814-7. doi: 10.1109/IEMBS.2005.1616801.

Authors

Wee Soon Yeoh¹, Cishen Zhang, Ming Chen

Affiliation

¹ Sch. of Electr. & Electron. Eng., Nanyang Technol. Univ.

PMID: 17282570
DOI: 10.1109/IEMBS.2005.1616801

Abstract

This paper uses a new tissue model, which incorporates random fluctuations of the tissue signal within the received ultrasound radio frequency (RF) echo signal, to propose an ultrasound image deconvolution algorithm for improving the quality of ultrasound images. Using this model, we propose a modified version of the classical regularization method which combines optimal Wiener filter and constrained least squares (LS) filtering algorithm for estimation of the tissue signal. The performance of the algorithm is evaluated by simulation. The deconvolved images can show significant improvement in both the speckle reduction and the contrast ratio.