人工智能影像分析软件对老年机械取栓患者手术效率及预后影响的单中心经验

Effect of AI software on surgical efficiency and prognosis in elderly patients with mechanical thrombectomy:single-center experience

目的探讨应用人工智能影像分析软件对急性大血管闭塞的老年脑卒中机械取栓患者手术效率及预后的影响。方法选择2019年9月至2022年11月苏州大学附属张家港医院收治的急性大血管闭塞机械取栓的老年患者80例,回顾性分析前瞻性登记的数据,根据是否应用人工智能影像分析软件分为人工智能组40例和对照组40例,对比2组手术时间、成功再灌注率、一把通率、90 d预后及死亡。结果2组入院至穿刺时间和术后成功再灌注率比较,差异无统计学意义(P>0.05)。人工智能组穿刺至再通时间、入院至再通时间明显低于对照组[58(40,80)min vs 70(60,80)min,188(167,216)min vs 246(170,288)min,P<0.05],一把通比例明显高于对照组(65.0%vs 35.0%,P<0.05)。多因素logistic回归分析显示,应用人工智能软件与患者90 d良好预后(改良的Rankin量表评分0~3分)、早期神经功能改善呈正相关(OR=9.737,95%CI:2.221~42.689,P=0.003;OR=3.683,95%CI:1.172~11.575,P=0.026),与病死率呈负相关(OR=0.166,95%CI:0.034~0.814,P=0.027)。结论急性大血管闭塞机械取栓患者术前应用人工智能影像分析软件有助于提高取栓手术效率,改善90 d预后和早期神经功能,并减少死亡等不良结局发生。

Objective To investigate the effect of artificial intelligence(AI)image analysis software on the efficacy of mechanical thrombectomy and prognosis in elderly patients with acute large vascular occlusive stroke.Methods A retrospective analysis was conducted on the prospectively recorded data of 80 elderly patients with acute large vessel occlusion treated with mechanical thrombectomy in our hospital from September 2019 to November 2022.According to the application of AI imaging analysis software or not,40 cases were divided into AI group and the other 40 into control group.Operative time,successful reperfusion rate,first-pass efficiency,90-day prognosis,and mortality were compared between the two groups.Results There were no statistical differences in the door to puncture time and the successful reperfusion rate between the two groups(P>0.05).The AI group had significantly shorter puncture to recanalization time and door to recanalization time[58(40,80)min vs 70(60,80)min,188(167,216)min vs 246(170,288)min,P<0.05],and obviously higer first-pass effect(65.0%vs 35.0%,P<0.05)than the control group.Multivariate logistic regression analysis showed that the use of AI software was positively correlated with 90-day good prognosis(mRS score 0-3)and early improvement of neurological function(OR=9.737,95%CI:2.221-42.689,P=0.003;OR=3.683,95%CI:1.172-11.575,P=0.026),and negatively with mortality(OR=0.166,95%CI:0.034-0.814,P=0.027).ConclusionFor patients with acute large vessel occlusion,application of AI image fast software before mechanical thrombectomy is helpful in promoting the efficiency of thrombectomy,improving the 90-day prognosis and early neurological function,and reducing the occurrence of adverse outcomes such as mortality.