摘要： 目的 对伴脑白质疏松症的超急性期高血压脑出血患者影像学资料进行回顾，评价血肿不稳定CT征象作为血肿扩大风险预测指标的可行性。方法 收集2021年3-4月在徐州医科大学附属医院急诊就诊、并于发病6 h内行头部CT平扫的95例高血压脑出血患者影像学资料，ITK?SNAP软件获得血肿体积，记录形态不规则征、混杂密度征、黑洞征、漩涡征、卫星征、岛征、液?液平征等血肿不稳定CT征象；Blennow量表评价脑白质疏松症严重程度；单因素和多因素逐步法Logistic回归分析筛查较大体积血肿（≥ 30 ml）危险因素。结果 根据血肿体积分为 < 30 ml组（53例）和≥ 30 ml组（42例），Logistic回归分析提示重度脑白质疏松症是高血压脑出血后发生较大体积血肿的危险因素（OR = 6.222，95%CI：2.506 ~ 15.452；P = 0.000）。血肿体积≥ 30 ml组血肿形态不规则征（χ² = 17.599，P = 0.000）、混杂密度征（χ² = 5.644，P = 0.018）、黑洞征（χ² = 23.161，P = 0.000）、漩涡征（χ² = 26.128，P = 0.000）、卫星征（χ² = 25.148，P = 0.000）和岛征（χ² = 20.575，P = 0.000）等单一征象，以及≥ 2种（χ² = 23.245，P = 0.000）和3种（χ² = 58.902，P = 0.000）联合征象比例均高于 < 30 ml组；重度脑白质疏松症（Ⅱ级和Ⅲ级）组（50例）血肿形态不规则征（χ² = 5.502，P = 0.019）、黑洞征（χ² = 10.715，P = 0.001）、漩涡征（χ² = 7.336，P = 0.007）、卫星征（χ² = 4.467，P = 0.035）和岛征（χ² = 19.961，P = 0.000），以及≥ 2种（χ² = 12.684，P = 0.000）和3种（χ² = 17.593，P = 0.000）联合征象比例均高于非重度（无病变和Ⅰ级）组（45例）。结论 重度脑白质疏松症是超急性期高血压脑出血血肿体积扩大（≥ 30 ml）的重要危险因素，发病急性期血肿不稳定CT征象阳性检出率较高，提示该指标可以作为脑出血后血肿扩大风险预测指标。

关键词: 颅内出血, 高血压性, 脑白质疏松症, 体层摄影术, 螺旋计算机, 危险因素, Logistic模型

Abstract: Objective To analyze the imaging data of patients with hyperacute hypertensive cerebral hemorrhage with leukoaraiosis (LA), and to evaluate the possibility of unstable CT signs of hematoma as a predictor of hematoma enlargement risk.Methods The imaging data of 95 patients who were admitted to the Department of Emergency of The Affiliated Hospital of Xuzhou Medical University from March to April, 2021 and underwent head CT scan within 6 h of onset were collected, and hematoma volume was obtained by ITK - SNAP software. Unstable CT signs of hematoma including irregular shape sign, mixed density sign, black hole sign, swirl sign, satellite sign, island sign and liquid-liquid level sign were recorded. The severity of LA was assessed by the Blennow scale. Univariate and multivariate stepwise Logistic regression analysis were used to screen the risk factors of large volume cerebral hemorrhage (≥ 30 ml).Results According to the hematoma volume of cerebral hemorrhage, 95 cases were divided into < 30 ml group (n = 53) and ≥ 30 ml group (n = 42). Logistic regression analysis showed that severe LA was a risk factor for large volume hematoma in cerebral hemorrhage (OR = 6.222, 95%CI: 2.506- 15.452; P = 0.000). CT showed that the proportions of irregular shape sign (χ² = 17.599, P = 0.000), mixed density sign (χ² = 5.644, P = 0.018), black hole sign (χ² = 23.161, P = 0.000), swirl sign (χ² = 26.128, P = 0.000), satellite sign (χ² = 25.148, P = 0.000), island sign (χ² = 20.575, P = 0.000), ≥ 2 combined signs (χ² = 23.245, P = 0.000) and ≥ 3 combined signs (χ² = 58.902, P = 0.000) were higher in the hematoma volume ≥ 30 ml group than those in < 30 ml group. Severe LA (grade Ⅱ and Ⅲ) group (n = 50) showed irregular shape sign (χ² = 5.502, P = 0.019), black hole sign (χ² = 10.715, P = 0.001), swirl sign (χ² = 7.336, P = 0.007), satellite sign (χ² = 4.467, P = 0.035), island sign (χ² = 19.961, P = 0.000), ≥ 2 combined signs (χ² = 12.684, P = 0.000) and ≥ 3 combined signs (χ² = 17.593, P = 0.000) were higher than those in non-severe (non and grade Ⅰ) group (n = 45).Conclusions Severe LA is an important risk factor for hematoma volume expansion (≥ 30 ml) in hyperacute hypertensive cerebral hemorrhage. The positive rates of unstable CT signs of hematoma in acute stage are higher, suggesting that this indicator can be used as a risk predictor for hematoma expansion after cerebral hemorrhage.

Key words: Intracranial hemorrhage, hypertensive, Leukoaraiosis, Tomography, spiral computed, Risk factors, Logistic models