Abstract:

Objective　To investigate the value of high-resolution magnetic resonance imaging (HR-MRI) image characteristics and quantitative parameters in predicting the risk of ischemic cerebral infarction, and analyze its influence on early clinical diagnosis. Methods　This protocol was designed according to the method of controlled trial, prospective, single-blind, and single-center clinical study. A total of 83 patients with transient ischemic attack admitted to Liaocheng Third People's Hospital from March 2020 to March 2022 were selected for HR-MRI examination. According to whether ischemic cerebral infarction occurred within one week, they were divided into an occurrence group [32 patients, 18 males and 14 females, aged (64.81±7.36) years] and a non-occurrence group [51 patients, 28 males and 23 females, aged (65.14±7.53) years]. The HR-MRI image characteristics [plaque distribution and plaque characteristics (plaque area and plaque load)], HR-MRI quantitative parameters [area of the narrowest part of blood vessel (VA_MLN), vascular area (VA_reference), wall area of the narrowest part of blood vessel (WA_MLN), lumen area of the narrowest part of blood vessel (LA_MLN), reference lumen area (LA_reference), and stenosis rate], and reconstruction modes (reconstruction index, positive reconstruction, negative reconstruction, and no significant reconstruction) of the two groups were compared. χ² test was used for plaque distribution and reconstruction mode, and t test for plaque characteristics and HR-MRI quantitative parameters. The receiver operating characteristic curve (ROC) was used to analyze the predictive value of HR-MRI image characteristics and quantitative parameters in the risk of ischemic cerebral infarction. Results　The plaque area [(6.43±3.06) mm²] and plaque load [(29.17±7.42)%] in the occurrence group were higher than those in the non-occurrence group [(5.37±1.02) mm² and (24.18±5.45)%], with statistically significant differences (t=2.287 and 3.525, both P<0.05). The positive reconstruction rate in the occurrence group [81.25% (26/32)] was higher than that in the non-occurrence group [52.94% (27/51)], with a statistically significant difference (χ²=6.827, P=0.009). The WA_MLN [(12.04±2.02) mm²] and stenosis rate [(32.46±2.02)%] in the occurrence group were higher than those in the non-occurrence group [(10.63±1.41) mm² and (21.20±1.35)%], while the LA_MLN [(10.22±1.75) mm²] was lower than that in the non-occurrence group [(11.96±2.07) mm²], with statistically significant differences (t=3.744, 30.462, and 3.949, all P<0.001). The areas under the curve (AUCs) of plaque area, plaque load, WA_MLN, LA_MLN, and stenosis rate for predicting ischemic cerebral infarction were 0.839 (95%CI 0.794-0.942), 0.706 (95%CI 0.598-0.820), 0.773 (95%CI 0.694-0.893), 0.819 (95%CI 0.809-0.894), and 0.812 (95%CI 0.731-0.910), all above 0.7. The AUC value [0.947 (95%CI 0.873-0.961)] of all indexes and parameters combined to predict the risk of ischemic cerebral infarction was the highest. Conclusion　HR-MRI plaque area, plaque load, WA_MLN, LA_MLN, and stenosis rate have high predictive values for the risk of ischemic cerebral infarction, which is helpful for the early diagnosis and intervention of ischemic cerebral infarction.

Key words:

Ischemic cerebral infarction, Transient ischemic attack, High resolution magnetic resonance imaging, Lumen area of the narrowest part of blood vessel, Reconstruction index

摘要：

目的　探讨高分辨率磁共振成像（HR-MRI）影像特征、定量参数对缺血性脑梗死发生风险的预测价值，分析其对临床早期诊断的影响。方法　本方案按对照试验、前瞻性、单盲、单中心临床研究方法设计。选取2020年3月至2022年3月聊城市第三人民医院收治的83例短暂性脑缺血发作患者，均行HR-MRI检查，根据1周内是否发生缺血性脑梗死分为发生组［32例，男18例、女14例，年龄（64.81±7.36）岁］、未发生组［51例，男28例、女23例，年龄（65.14±7.53）岁］。对比两组HR-MRI影像特征［斑块分布、斑块特征（斑块面积、斑块负荷）］、HR-MRI定量参数［血管最狭窄处面积（VA_MLN）、血管面积（VA_reference）、血管最狭窄处管壁面积（WA_MLN）、血管最狭窄处管腔面积（LA_MLN）、参考血管腔面积（LA_reference）及狭窄率］、重构方式（重构指数、正性重构、负性重构、无明显重构）。斑块分布、重构方式以χ²检验，斑块特征、HR-MRI定量参数以t检验；采用受试者工作特征曲线（ROC）分析HR-MRI影像特征、定量参数对缺血性脑梗死发生风险的预测价值。结果　发生组斑块面积［（6.43±3.06）mm²］、斑块负荷［（29.17±7.42）%］均大于未发生组［（5.37±1.02）mm²、（24.18±5.45）%］，差异均有统计学意义（t=2.287、3.525，均P<0.05）。发生组正性重构率［81.25%（26/32）］高于未发生组［52.94%（27/51）］，差异有统计学意义（χ²=6.827，P=0.009）。发生组WA_MLN［（12.04±2.02）mm²］、狭窄率［（32.46±2.02）%］均高于未发生组［（10.63±1.41）mm²、（21.20±1.35）%］，LA_MLN［（10.22±1.75）mm²］低于未发生组［（11.96±2.07）mm²］，差异均有统计学意义（t=3.744、30.462、3.949，均P<0.001）。斑块面积、斑块负荷、WA_MLN、LA_MLN、狭窄率预测缺血性脑梗死发生的曲线下面积（AUC）分别为0.839（95%CI 0.794～0.942）、0.706（95%CI 0.598～0.820）、0.773（95%CI 0.694～0.893）、0.819（95%CI 0.809～0.894）、0.812（95%CI 0.731～0.910），均在0.7以上，且各项指标、参数联合预测缺血性脑梗死发生风险的AUC值［0.947（95%CI 0.873～0.961）］最高。结论　HR-MRI斑块面积、斑块负荷、WA_MLN、LA_MLN及狭窄率对缺血性脑梗死发生风险具有较高预测价值，有利于临床对缺血性脑梗死的早期诊断、干预。

关键词:

缺血性脑梗死, 短暂性脑缺血发作, 高分辨率磁共振成像, 血管最狭窄处管腔面积, 重构指数