Kidney stone specific ultrasound imaging of human subjects

Abstract

Sensitivity and specificity data show users have more difficulty in identifying stones accurately in B-mode ultrasound than x-ray CT. Our goal was to evaluate the signal to noise (SNR) of a new stone specific imaging algorithm, S-mode based on Color Doppler twinkling artifact, to B-mode. Forty sets of B- and S-mode imaging data were collected from 16 subjects using a Philips HDI C5-2 imaging probe and Verasonics ultrasound system. Two ways S-mode differs from Doppler is that it filters out blood flow signal and uses reverse color write priority to add color to echogenic regions only. For both B- and S-mode raw data, we calculate SNR of the magnitude (brightness) of the stone signal compared to the second highest magnitude in the image. The mean and standard deviation of the SNR was 1.6 ± 0.7 for B-mode and 37 ± 24 for S-mode, with 1 being the stone is equally bright as, and difficult to distinguish from, background. In this human study of S-mode, stones appeared over 30 times brighter than background and with over 20 times the contrast to background seen in B-mode. [Work supported by NIH NIDDK grants DK043881 and DK092197, and NSBRI through NASA NCC 9-58.]