BRAIN评分联合影像组学的诺莫图预测自发性脑出血早期血肿扩大  

Nomogram of BRAIN Score Combined with Radiomics Predicts Early Expansion of Spontaneous Intracerebral Hemorrhage

作　　者：宋岩 于亚超 荣梦露 张红娟[1] 张跃跃[3] 魏超刚 沈钧康[3] 肖岳[2] SONG Yan;YU Yachao;RONG Menglu(Department of Radiology,Jieshou City People’s Hospital/Jieshou Hospital Affiliated to Anhui Medical College,Fuyang,Anhui Province 236500,P.R.China)

机构地区：[1]界首市人民医院/安徽医学高等专科学校附属界首医院医学影像科,236500 [2]界首市人民医院/安徽医学高等专科学校附属界首医院重症医学科,236500 [3]苏州大学附属第二医院医学影像科,215004

出　　处：《临床放射学杂志》2025年第3期403-410,共8页Journal of Clinical Radiology

基　　金：国家自然科学青年基金项目(编号:81801754);苏州市“科教兴卫”青年科技项目(编号:KJXW2023022);安徽卫生健康科研项目(编号:AHWJ2024Ab0069)。

摘　　要：目的探究自发性脑出血(sICH)患者早期血肿扩大(EHE)的因素,并构建基于BRAIN评分量表联合影像组学的诺莫图预测模型。方法123例sICH患者,按照7∶3的比例分为训练集(n=86)与验证集(n=37)。搜集患者的临床及CT特征资料,采用Logistic回归筛选EHE的危险因素,利用Darwin平台提取并筛选影像组学特征,建立CT影像组学预测模型、CT征象预测模型、BRAIN评分预测模型及三者联合的预测模型,并对其进行验证。结果EHE组与非EHE组在BRAIN评分、混杂密度征、岛征、黑洞征、液平征、漩涡征等因素存在统计学差异(P<0.05);多因素Logistic回归分析确认这些因素为EHE的危险因素(P<0.05)。构建的联合模型预测效能最高,其训练集和验证集的受试者工作特征(ROC)曲线下面积分别为0.924和0.871,校正曲线显示预测值与实际值一致性良好。决策曲线分析显示,列线图模型在多数阈值概率下具有较高的净获益值。结论基于BRAIN评分联合影像组学的列线图模型在预测sICH患者EHE方面具有较高的准确性,可为临床诊断和治疗决策提供参考依据。Objective To explore the risk factors for early hematoma expansion(EHE)in patients with spontaneous intracerebral hemorrhage(sICH),and to construct a nomogram prediction model based on BRAIN score combined with radiomics.Methods A total of 123 sICH patients were divided into a training set(n=86)and a validation set(n=37)in a ratio of 7∶3.Clinical and CT characteristic data of patients were collected,and the risk factors of EHE were screened by logistic regression.The imaging features were extracted and screened by the Darwin platform,and prediction models of CT radiomics,CT signs,BRAIN score,and their combination were established and validated.Results There were significant differences in BRAIN score,mixed density sign,island sign,black hole sign,liquid level sign,and vortex sign between the EHE group and the non-EHE group(P<0.05).Multivariate logistic regression analysis confirmed these factors as risk factors for EHE(P<0.05).The combined model had the highest prediction efficiency,with the area under the ROC curve of the training set and the validation set being 0.924 and 0.871,respectively.The calibration curve showed that the predicted value was in good agreement with the actual value.Decision curve analysis shows that the nomogram model has a higher net benefit value under most threshold probabilities.Conclusion The nomogram model based on BRAIN score combined with radiomics has high accuracy in predicting EHE in sICH patients,which can provide reference for clinical diagnosis and treatment decisions.

关 键 词：自发性脑出血 早期血肿扩大 BRAIN评分 影像组学 列线图模型 

分 类 号：R743.34[医药卫生—神经病学与精神病学] R816.1[医药卫生—临床医学]