| 张爱琴,王春利,边桂然,等.基于振幅整合脑电图和乳酸的晚期早产儿脑损伤早期预测模型构建分析[J].安徽医药,2025,29(6):1156-1160. |
| 基于振幅整合脑电图和乳酸的晚期早产儿脑损伤早期预测模型构建分析 |
| Construction and analysis of an early prediction model for brain injury in late preterm infants based on amplitude-integrated electroencephalogram and lactate |
| |
| DOI：10.3969/j.issn.1009-6469.2025.06.018 |
| 中文关键词: 脑损伤 脑皮层电图 振幅整合脑电图 乳酸 预测模型 婴儿，早产 |
| 英文关键词: Brain injury Electrocorticography Amplitude integrated EEG Lactic acid Prediction model Infant, premature |
| 基金项目:沧州市重点研发计划指导项目( 213106051) |
|
| 摘要点击次数: 48 |
| 全文下载次数: 22 |
| 中文摘要: |
| 目的基于振幅整合脑电图( aEEG)和乳酸建立晚期早产儿脑损伤早期预测模型，并分析预测模型预测效能。方法选取 2019年 9月至 2021年 8月沧州市人民医院出生并立即转入新生儿科住院的晚期早产儿(胎龄 34~36+6周) 200例作为建模组，于早产儿出生后 1h内进行动脉血乳酸测定，在早产儿出生后 72 h内行 aEEG监测，并评估病儿出生后 4~14 d脑损伤发生情况，根据脑损伤情况分为脑损伤组和无脑损伤组，利用 logistic回归分析晚期早产儿脑损伤的危险因素，建立预测模型，并分析预测模型预测效能；另纳入同期该院 100例晚期早产儿作为验证组以验证模型效能。结果 200例晚期早产儿在出生后的 4~14 d内共 38例病儿发生脑损伤(脑损伤组)162例病儿未发生脑损伤(无脑损伤组)。建模组 logistic分析显示:机械通气治疗( OR=1.59)、高乳酸水平( OR=1.25)是晚期早产，儿脑损伤发生的危险因素( P<0.05)高 1 min Apgar评分( OR=0.51)、高 aEEG评分( OR=0.55)是晚期早产儿脑损伤发生的保护因素( P<0.05)；建模组、验证组受试者操，作特征曲线( ROC曲线)分析显示:基于 aEEG、1 min Apgar评分、机械通气治疗、乳酸水平构建的预测模型预测晚期早产儿脑损伤的曲线下面积( AUC)分别为 0.94、0.91，表明基于 aEEG和乳酸水平构建的晚期早产儿脑损伤早期预测模型预测效能较高。结论基于 aEEG和乳酸建立的晚期早产儿脑损伤早期预测模型可用于晚期早产儿脑损伤早期预测。 |
| 英文摘要: |
| Objective To establish an early prediction model for brain injury in late preterm infants based on amplitude-integrated electroencephalogram (aEEG) and lactate levels, and to evaluate the model's predictive performance.Methods A total of 200 late pre‐term infants (gestational age: 34-36+6 weeks) born at Cangzhou People's Hospital from September 2019 to August 2021 and immediatelyadmitted to the Department of Neonatology were enrolled as the modeling group. Arterial blood lactate was measured within 1 hour afterbirth, and aEEG monitoring was performed within 72 hours. Brain injury was assessed between 4-14 days postpartum, with infants clas-sified into brain injury and non-brain injury groups. Logistic regression identified risk factors for brain injury and established the pre-diction model. An additional 100 late preterm infants from the same hospital were included as a validation cohort.Results Among the 200 infants, 38 developed brain injury (brain injury group) and 162 did not (non-brain injury group). Logistic regression analysis re-vealed mechanical ventilation (OR=1.59) and elevated lactate levels (OR=1.25) as significant risk factors (P<0.05), while higher 1-min-ute Apgar scores (OR=0.51) and aEEG scores (OR=0.55) were protective factors (P<0.05). Receiver operating characteristic (ROC)curve analysis for both the modeling and validation groups revealed that the area under the curve (AUC) of the prediction model basedon aEEG, 1-minute Apgar score, mechanical ventilation, and lactate level was 0.94 and 0.91, respectively. This suggests that the earlyprediction model for brain injury in late preterm infants based on aEEG and lactate level has high predictive efficiency.Conclusion The early prediction model for brain injury in late preterm infants based on aEEG and lactate can be used effectively used for early pre-diction of brain injury in this population. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
