Update on carotid plaque instability quantification using strain indices from multiple regions of interest in carotid plaque

Abstract

Carotid plaque rupture is a primary cause of ischemic strokes and transient ischemic attacks (TIAs). Probability of stroke and TIAs depend on the mechanical stability of plaque. Ultrasound based Lagrangian carotid strain imaging (LCSI) can provide a non-invasive assessment of plaque stiffness to assess mechanical stability. We report on ultrasound strain indices from multiple regions of interest (ROI) in plaque as biomarkers for plaque instability. Our study included 70 patients who participated in pre-surgical ultrasound LCSI and cognitive testing. Patients were classified as symptomatic or asymptomatic based on reported clinical symptoms. We hypothesize that patients with unstable plaque are more likely to have cognitive deficits due to undetected ischemic events. Previous work evaluated strain indices from a single ROI. This work extends analysis into a multiple ROI approach. Five strain indices are identified as potentially useful biomarkers of plaque instability. These include the maximum accumulated strain indices (MASI) in the primary ROI used previously, mean of the standard deviation of strain estimated in entire plaque (MSD), the average L1 Norm (AL1NWP), total number of unstable ROIs (NUROI), and the percent of unstable estimates in the entire plaque (PPS). Using four indices with highest correlations resulted in an area under curve (AUC) value of 0.80. However, since NUROI and PPS are independent and not directly related to strain a combination of the two top performing indices with NUROI and PPS provided an AUC of 0.85. Our multiple ROI approach thus demonstrated independent and improved indices for plaque instability quantification.