Automated intima-media thickness: the Los Angeles Atherosclerosis Study

Common carotid artery (CCA) intima-media thickness (IMT) from B-mode ultrasound is a widely used measure of early atherosclerosis. This study evaluated within- and between-sonographer reproducibility of automated edge-tracking IMT using a low-cost mobile scanner. B-mode images of the left and right CCA were acquired on two occasions (interval of 2-14 days) by two sonographers for 38 subjects, aged 31-75 y. Reproduciblity error was measured as the mean absolute difference (MAD+/-SEM) and the standard deviation of differences (SDdelta) between repeated measurements. Within-sonographer (MAD=0.027+/-0.006 mm; SDdelta=0.044 mm) and between-sonographer errors (MAD=0.041+/-0.008 mm; SDdelta=0.064) in IMT (mean=0.74, SD=0.14) of a single artery were small compared to those of other protocols. Combined averaging across both body positions and arteries reduced intersonographer MAD by 47% (MAD=0.022+/-0.003 mm; SDdelta=0.029 mm). These data demonstrate that the proposed IMT protocol reduces reproducibility error by more than 50% relative to several protocols used in other major studies.