RESUMO
Purpose: The present study aimed to explore the predictive ability of an ultrasound linear regression equation in patients undergoing endovascular stent placement (ESP) to treat carotid artery stenosis-induced ischemic stroke. Methods: Pearson's correlation coefficient of actual improvement rate (IR) and 10 preoperative ultrasound indices in the carotid arteries of 64 patients who underwent ESP were retrospectively analyzed. A predictive ultrasound model for the fitted IR after ESP was established. Results: Of the 10 preoperative ultrasound indices, peak systolic velocity (PSV) at stenosis was strongly correlated with postoperative actual IR (r = 0.622; P < 0.01). The unstable plaque index (UPI; r = 0.447), peak eccentricity ratio (r = 0.431), and plaque stiffness index (ß; r = 0.512) moderately correlated with actual IR (P < 0.01). Furthermore, the resistance index (r = 0.325) and the dilation coefficient (r = 0.311) weakly correlated with actual IR (P < 0.05). There was no significant correlation between actual IR and the number of unstable plaques, area narrowing, pulsatility index, and compliance coefficient. In combination, morphological, hemodynamic, and physiological ultrasound indices can predict 62.39% of neurological deficits after ESP: fitted IR = 0.9816 - 0.1293ß + 0.0504UPI - 0.1137PSV. Conclusion: Certain carotid ultrasound indices correlate with ESP outcomes. The multi-index predictive model can be used to evaluate the effects of ESP before surgery.
