目的：探讨中性粒细胞/淋巴细胞比值（neutrophil-to-lymphocyte ratio，NLR）联合血肿影像 学 特 征 对 自 发 性 基 底 节 区 脑 出 血 后 发 生 迟 发 性 血 肿 周 围 脑 水 肿（delayed perihematomal edema，dPHE）的预测价值。

方法：回顾性纳入 2017年 1月—2023年 12月浙江大学医学院附属金华医院（金华市中心 医院）神经内科收治的自发性基底节区出血患者，采用单因素分析和多因素二元logistic向 前似然比（likelihood ratio，LR）偏似然估计法回归分析比较 dPHE 组和非 dPHE 组临床资 料 ，确 定 dPHE 的 独 立 影 响 因 素 ，应 用 受 试 者 工 作 特 征 曲 线（receiver operating characteristic，ROC）评估该模型对dPHE的预测价值。

结果：共纳入患者372例，其中dPHE组115例，非dPHE组257例。单因素分析显示，dPHE 组基线美国国立卫生研究院脑卒中量表（National Institutes of Health Stroke Scale， NIHSS）评分、基线血肿体积、超敏C反应蛋白、淋巴细胞数、NLR、混杂征、外侧型血肿、甘露 醇大剂量比例均显著高于非dPHE组（P<0.05）；且dPHE组出院时NIHSS评分、住院费用也 高于非 dPHE 组（P<0.05），其住院时间也更长（P<0.05）。进一步行多因素分析发现，NLR ［比值比（odds ratio，OR）＝1.138（95%CI：1.056～1.227）；P=0.001］、基线血肿体积［OR＝ 1.187（95% CI：1.132～1.246）；P<0.001］、外 侧 型 血 肿［OR＝3.067（95% CI：1.611～ 5.838）；P=0.001］是 dPHE 的独立预测因素。ROC 曲线分析显示，该模型的曲线下面积为 0.810（95% CI：0.765～0.856）。

结论：基线NLR越高，血肿体积越大，外侧型的基底节区出血患者更易发生dPHE。

Objective: To investigate the predictive value of neutrophil to lymphocyte ratio (NLR) combined with hematoma imaging features for delayed perihematoma edema (dPHE) after spontaneous basal ganglia hemorrhage.

Methods: Retrospective inclusion of patients with spontaneous basal ganglia hemorrhage admitted to the neurology department of Jinhua Hospital affiliated to Zhejiang University School of Medicine (Jinhua Municipal Central Hospital) from January 2017 to December 2023. We compared clinical data between the dPHE group and no-dPHE group based on the univariate analysis and multivariate binary logistic regression analysis. We identified the independent influencing factors of dPHE, and evaluated the predictive value of the model for dPHE using receiver operating characteristic (ROC).

Results: A total of 372 patients were included in this study, including 115 in the dPHE group and 257 in the no-dPHE group. Univariate analysis showed that the baseline National Institutes of Health Stroke Scale (NIHSS) score, baseline hematoma volume, high-sensitivity Creactive protein, lymphocyte count, NLR, blend sign, lateral hematoma, and high-dose mannitol proportion in the dPHE group were significantly higher than those in the non-dPHE group (P<0.05). Moreover, the NIHSS score and hospitalization expenses in the dPHE group were higher than those in the no-dPHE group (P<0.05), and their hospitalization time was also longer in the dPHE group (P<0.05). Furthermore, the multivariate logistic regression analysis revealed that NLR [odds ratio (OR) ＝1.138 (95% CI: 1.056-1.227); P＝0.001], baseline hematoma volume [OR＝1.187 (95% CI: 1.132-1.246); P<0.001] and lateral hematoma [OR＝3.067 (95% CI: 1.611-5.838); P＝0.001] were independent predictive factors of dPHE. ROC analysis showed that the area under the curve of the model was 0.810 (95% CI: 0.765-0.856).

Conclusion: The higher the baseline NLR, the larger the hematoma volume, and patients with lateral basal ganglia hemorrhage are more susceptible to dPHE.