TABLE 1. An Overview of Texture Feature Variability Studies in Ultrasound Video of the Atherosclerotic Carotid Plaque.

Study	Year	Method	NR (N/A/S)	Findings
Kanber [8]	2013	Manual segmentation	-/-/47	Carotid artery diameter not correlated with GSM over the cardiac cycle. No statistical significance between systole and diastole for plaque GSM (p=0.34). GSM correlated with presence of symptoms (p=0.02).
Kanber [9]	2013	Manual segmentation	-/-/27	Neither systolic nor diastolic states were estimated for studying GSM variability. Periodic variations of GSM were recorded as follows: GSM 40.5±1.69 (%CV=4.2%) vs 34.8±1.83 (%CV=5.2%).
Golemati [4]	2015	Manual segmentation	-/25/-	Found SD differences between wall and plaque for low stenosis subjects for contrast (8.56±7.98 vs 8.88±9.1 at wall) and correlation (0.07±0.05 vs 0.04±0.16 at plaque) at systole vs diastole respectively.
This study	2016	Semi-automated segmentation based on snakes	40/40/40	TF may separate N, A and S groups for both the cardiac systolic and diastolic states. For example, GSM at systole vs diastole were: 42.12±3.15 vs 39.38±2.01, p=0.001 for the N group, 44.10±1.84 vs 34.88±1.96, p=0.001 for the A group and 35.73±1.69 vs 29.01±2.25, p=0.001 for the S group.

CVD: Cardiovascular events; N: Number of cases investigated; significantly (shown underlined) and non-significantly different at p<0.05 and p>=0.05 respectively.