摘要
目的探讨基于表观扩散系数(apparent diffusion coefficient,ADC)最小值(ADCmin)及临床和影像特征构建模型预测高强度聚焦超声(high intensity focused ultrasound,HIFU)治疗子宫肌瘤的临床疗效。材料与方法回顾性收集2021年9月至2023年12月行子宫肌瘤HIFU治疗并符合纳排标准的153例患者临床及影像资料,共153枚肌瘤纳入研究(单病例存在多个肌瘤取其肌瘤最大枚)。超声评估HIFU治疗3个月后肌瘤体积缩小率,并将其分为疗效显著组(≥50%体积缩小,n=62)和非显效组(<50%体积缩小,n=91)。测量子宫肌瘤ADCmin和ADC平均值(ADC_(mean)),采用单因素分析和多因素logistic回归对临床和影像资料进行筛选后建立临床影像模型。通过受试者工作特征(receiver operating characteristic,ROC)曲线比较子宫肌瘤ADCmin和ADC_(mean)的预测效能,选择子宫肌瘤ADC具有较高预测效能的量化指标与临床及影像特征结合构建联合模型。ROC评估ADCmin、临床影像模型及联合模型的预测效能,DeLong检验比较不同模型间ROC曲线下面积(area under the curve,AUC)的差异。结果共筛出血红蛋白、身体质量指数(body mass index,BMI)、T1WI增强信号程度三个因素建立临床影像特征模型。ROC曲线显示ADCmin、ADC_(mean)的AUC值分别为0.753[95%置信区间(confidence interval,CI):0.677~0.828]、0.658(95%CI:0.570~0.746),DeLong检验结果表明ADCmin的预测效能高于ADC_(mean)(P<0.05)。临床影像模型及ADCmin与临床影像特征联合模型的AUC值分别为0.711(95%CI:0.627~0.796)、0.816(95%CI:0.748~0.884),DeLong检验显示ADCmin的预测效能与临床影像模型的AUC差异无统计学意义(P>0.05)。ADCmin与临床及影像特征联合模型的预测效能优于ADCmin值(P<0.05)及临床影像特征模型(P<0.05)。结论ADCmin与临床及影像特征构建的联合模型可有效预测HIFU治疗子宫肌瘤的临床疗效,并能为临床治疗方案的制订提供一定参考依据。
Objective:Construct a model to predict the clinical efficacy of high intensity focused ultrasound(HIFU)in treating uterine fibroids based on the apparent diffusion coefficient minimum(ADCmin)and clinical and imaging features.Materials and Methods:Clinical and imaging data of 153 patients who underwent HIFU treatment for uterine fibroids from September 2021 to December 2023 and met the criteria for inclusion were retrospectively collected,and a total of 153 fibroids were included in the study(choosing the largest in cases with multiple).Ultrasound assessed the fibroids'volume reduction rate three months after HIFU treatment,categorizing them into a significantly effective group(≥50%volume reduction,n=62)and a non-effective group(<50%volume reduction,n=91).Measurements were taken for the uterine fibroids'ADC_(min) and apparent diffusion coefficient mean(ADC_(mean)).Univariate and multivariate logistic regression analyses were performed on clinical and imaging data,identifying factors to establish a clinical imaging feature model.Receiver operating characteristic(ROC)curves compared the predictive performance of uterine fibroids'ADC_(min) and ADC_(mean).A combined model was constructed by integrating the quantifiable indicator with high predictive efficiency,ADCmin,and the clinical imaging feature model.ROC evaluation and DeLong test compared the area under the curve(AUC)differences between ADCmin,the clinical imaging feature model,and the combined model.Results:Hemoglobin,body mass index(BMI),and T1WI enhancement signal intensity were identified as factors to establish a clinical imaging feature model.The ROC curve demonstrates that the AUC values for ADC_(min) and ADC_(mean) were 0.753[95%confidence interval(CI):0.677-0.828]and 0.658(95%CI:0.570-0.746),respectively.The DeLong test results show that ADC_(min) demonstrates higher predictive performance than ADC_(mean)(P<0.05).The AUC values for the clinical imaging feature model and the combined ADC_(min) and clinical imaging feature model were 0.711(95%CI:0.627-0.796)and 0.816(95%CI:0.748-0.884),respectively.DeLong test indicates that the difference in AUC between ADC_(min) and the clinical imaging model is not statistically significant(P>0.05).The predictive efficacy of the combined model of ADC_(min) and clinical imaging features surpasses that of ADC_(min) alone(P<0.05)and the clinical imaging feature model(P<0.05).Conclusions:The combined model constructed by ADCminand clinical and imaging features can effectively predict the clinical efficacy of HIFU in the treatment of uterine fibroids,and can offer valuable insights for formulating clinical treatment strategies.
作者
刘姿延
黄小华
刘子熠
王渊
万夕瑶
蒋雨
LIU Ziyan;HUANG Xiaohua;LIU Ziyi;WANG Yuan;WAN Xiyao;JIANG Yu(Department of Radiology,Affiliated Hospital of North Sichuan Medical College,Nanchong 637000,China)
出处
《磁共振成像》
CAS
CSCD
北大核心
2024年第9期101-106,共6页
Chinese Journal of Magnetic Resonance Imaging
基金
南充市市校合作项目(编号:19SXHZ0429)。
关键词
子宫肌瘤
高强度聚焦超声
磁共振成像
表观扩散系数
预测
uterine fibroids
high intensity focused ultrasound
magnetic resonance imaging
apparent diffusion coefficient
prediction
