Prognostic Prediction Model for Endovascular Treatment of Acute Anterior Circulation Ischemic Stroke Based on Rapid Processing of PerfusIon and Diffusion Software.

Abstract

To construct a universally applicable nomogram using computed tomography perfusion parameters based on Rapid processing of PerfusIon and Diffusion software for evaluating the prognosis of acute anterior circulation large vessel occlusion treated by endovascular therapy. A total of 130 patients were divided into 2 groups: a group with a good prognosis [Rankin Rating Scale (mRS) score ≤2 points] and a group with a poor prognosis (mRS score >2 points). The least absolute shrinkage and selection operator logistic regression method was used to screen the predictive factors, and a nomogram chart prediction model was established. There were 52 patients in the good prognosis group and 78 patients in the poor prognosis group. Logistic regression analysis showed that age, baseline National Institute of Health Stroke Scale score, and volumes of cerebral blood flow < 30% were independent influencing factors for the prognosis of patients [odds ratio (OR = 1.065, 95% confidence interval (CI): 1.007-1.126; OR = 1.097, 95% CI: 1.008-1.193; OR = 1.107, 95% CI: 1.024-1.197; P all < 0.05]. Our nomogram model showed good discrimination and calibration. The receiver operating characteristic curve analysis showed that the area under the curve of the prediction model was 0.809 (95% CI: 0.733-0.885, P < 0.001), and the calibration curve was close to the ideal curve. Our nomogram can be used as a simple and reasonable tool to predict the prognosis of acute anterior circulation large vessel occlusion treated by endovascular therapy. The computed tomography perfusion parameter volumes of cerebral blood flow < 30% obtained by Rapid processing of PerfusIon and Diffusion software was an important predictive factor in this model.