通讯作者: 冯艳红, 13704067100@163.com

DOI：10.12201/bmr.202512.00047

• 摘要：【】目的 本研究探讨超声自动应变技术参数在射血分数保留心力衰竭(HFpEF)诊断中的价值,并构建一个融合临床指标、常规超声参数及自动应变参数的列线图诊断模型,提升HFpEF早期诊断准确性。方法 选取2024年3月至2025年6月锦州医科大学附属第一医院就诊的313例患者,其中216例经H2FPEF、HFA-PEFF高评分确诊的HFpEF患者作为实验组,97例存在危险因素的非HFpEF患者作为对照组。收集所有患者一般临床资料、常规超声心动图参数,及自动应变技术测得的左心室整体纵向应变(GLS)等参数。结果 通过单因素与多因素Logistic回归确定独立影响因素并构建列线图模型,再经ROC曲线、校准曲线、决策曲线及临床影响曲线分析模型性能,结果显示最终模型纳入7个变量:NT-proBNP(OR=19.133,95% CI= 10.119-36.179,P＜0.001)、E/e′比值(OR=5.533,95% CI=3.292-9.301,P＜0.001)、LVGLS(OR= 4.504,95% CI=2.552-7.949,P＜0.001)、LVAP2LS(OR=3.431,95% CI=2.074-5.677,P＜0.001)、LVAP3LS(OR=4.602,95% CI=2.762-7.67,P＜0.001)、LVAP4LS(OR=5.505,95% CI=2.944-10.295,P＜0.001)和LASr(OR=3.041,95% CI=1.853-4.993,P＜0.001)。训练集 ROC 曲线下面积(AUC)为 0.972(95% CI=0.954–0.989),测试集 AUC为 0.967(95% CI=0.936–0.997)。Hosmer-Lemeshow检验表明,预测与实际临床概率无统计学差异(训练集P=0.933,测试集P=0.766),校准曲线亦证实模型拟合优度良好；决策曲线分析显示,模型在全阈值概率下可获净收益,提示其具备良好临床效用。结论 基于超声自动应变技术核心参数,结合临床及常规超声指标构建的列线图诊断模型,对HFpEF具有良好的诊断效能,可作为辅助临床医生进行HFpEF个体化诊断的有效可视化工具,提高早期识别率,进而改善患者的管理和预后。

  关键词： ; ; 

   

  Abstract: Objective To explore the value of ultrasound automatic strain technique parameters in the diagnosis of heart failure with preserved ejection fraction (HFpEF), and to construct a nomogram diagnostic model integrating clinical indicators, conventional ultrasound parameters and automatic strain parameters, in order to improve the accuracy of early diagnosis of HFpEF.Methods Jinzhou Medical University from March 2024 to June 2025 were selected. Among them, 216 patients diagnosed with heart failure with preserved ejection fraction (HFpEF) based on high scores of H2FPEF and HFA-PEFF were included in the experimental group, while 97 patients with risk factors but without HFpEF were assigned to the control group. All patients general clinical data, conventional echocardiographic parameters, and parameters such as global longitudinal strain (GLS) of the left ventricle obtained by automatic strain technology were collected.Results Univariate and multivariate Logistic regression were used to determine independent influencing factors and construct a nomogram model. On this basis, the model performance was evaluated by drawing ROC curves, calibration curves, decision curves, and clinical impact curves.Results: The final model includes 7 variables: NT-proBNP (OR = 19.133, 95% CI = 10.119 - 36.179, P < 0.001), E/e ratio (OR = 5.533, 95% CI = 3.292 - 9.301, P < 0.001), LVGLS (OR = 4.504, 95% CI = 2.552 - 7.949, P < 0.001), LVAP2LS (OR = 3.431, 95% CI = 2.074 - 5.677, P < 0.001), LVAP3LS (OR = 4.602, 95% CI = 2.762 - 7.67, P < 0.001), LVAP4LS (OR = 5.505, 95% CI = 2.944 - 10.295, P < 0.001), and LASr (OR = 3.041, 95% CI = 1.853 - 4.993, P < 0.001). The area under the ROC curve (AUC) of the training set was 0.972 (95% CI = 0.954 - 0.989), and the AUC of the test set was 0.967 (95% CI = 0.936 - 0.997). The Hosmer-Lemeshow test showed no statistical difference between the predicted and actual clinical probabilities (training set P = 0.933, test set P = 0.766), and the calibration curve also indicated a good fit of the model. Additionally, the decision curve analysis revealed that the model could achieve net benefits across the entire threshold probability range, indicating that the model has good clinical utility.Conclusion Based on the core parameters of the ultrasonic automatic strain technology, and combined with clinical and conventional ultrasound indicators, the diagnostic model constructed by the nomogram has good diagnostic efficacy for HFpEF. It can be used as an effective visual tool to assist clinicians in individualized diagnosis of HFpEF, improve the early recognition rate, and thereby enhance the management and prognosis of patients.

  Key words: Ejection fraction retention heart failure; Automatic strain technology; Prediction model: Nomogram

  提交时间：2025-12-16

  版权声明：作者本人独立拥有该论文的版权，预印本系统仅拥有论文的永久保存权利。任何人未经允许不得重复使用。
• html

• 图表

• 周航宇, 李杰. 射血分数保留的心力衰竭合并慢性肾脏疾病的研究进展. 2025. doi: 10.12201/bmr.202504.00077

  蔡杰, 韩陈媛. 某三级医院慢性射血分数降低心力衰竭住院患者指南导向药物应用分析与评价. 2025. doi: 10.12201/bmr.202507.00076

  胡欣. 剖宫产术中寒战风险预测模型及列线图的构建. 2025. doi: 10.12201/bmr.202501.00053

  李明峻, 翟宁. 人工智能在心力衰竭精准分型中的研究进展与展望. 2025. doi: 10.12201/bmr.202512.00030

  张二利, 何兰兰, 李丹阳, 沈立, 吴钟华, 张军, 叶永强. 基于脑小血管病影像学总负荷的急性脑梗死患者住院时间延长列线图模型构建与验证. 2025. doi: 10.12201/bmr.202501.00008

  蔡欣欣, 吴夏阳. 早产儿入院败血症风险预测模型的构建与验证. 2025. doi: 10.12201/bmr.202505.00013

  阮旭凌, 刘琦, 郭志恒, 晏峻峰. 基于LDA和XGBoost算法的乳腺癌预测模型构建研究. 2022. doi: 10.12201/bmr.202106.00007

  温婉婷婷, 谢慰, 王若宇, 彭建平, 张涛, 陈斌, 伍玉南. 吲哚菁绿15分钟滞留率联合血浆氨、MELD评分对HBV相关慢加急性肝衰竭患者预后预测价值. 2025. doi: 10.12201/bmr.202507.00064

  贾李蓉. 中医药自动问答系统的问题理解研究. 2021. doi: 10.12201/bmr.202101.00002

  付思思. 非小细胞肺癌患者肺叶切除术后恶心呕吐Nomogram分析模型构建与验证. 2025. doi: 10.12201/bmr.202507.00001

• 序号	提交日期	编号	操作
  1	2025-11-08	10.12201/bmr.202512.00047V1	下载

• 公开评论  匿名评论  仅发给作者 提交评论