左室节段应变识别小鼠内源与外源应激后的独特心肌变形模式
研究将6周龄雄性C57BL/6J小鼠随机分为瘦组与肥胖组(每组n=24),肥胖组接受12周高脂饮食诱导肥胖(DIO),18周时再随机分组接受8周不可预测慢性温和应激(UCMS)。于基线、高脂饮食后及UCMS后行超声心动图,应用机器学习分析节段应变。结果:区分肥胖与瘦鼠的AUC=0.921,侧壁及前游离壁径向应变变化最具信息量;预测UCMS暴露(不论饮食)的AUC=0.886,前壁中层纵向应变率与后间隔壁径向应变为首要特征。
要点
- 节段应变可区分DIO与UCMS两种应激,变形改变分别好发于游离壁与间隔壁。
- 肥胖与瘦鼠分类AUC=0.921,侧壁及前游离壁径向应变贡献最大。
- UCMS预测AUC=0.886,前壁中层纵向应变率与后间隔壁径向应变为关键特征。
关键影像表现/征象
- 肥胖与瘦鼠比较,侧壁径向应变变化-64%(p≤0.001)和前游离壁径向应变变化-53%(p<0.001)最具信息量。
- 预测UCMS暴露(不论饮食)的首要特征为前壁中层纵向应变率和后间隔壁径向应变。
- 变形模式改变好发于游离壁(DIO)与间隔壁(UCMS),提示疾病特异性心肌改变。
诊断标准与关键数值
- 区分肥胖与瘦鼠的机器学习模型AUC=0.921。
- 预测UCMS暴露(不论饮食)的机器学习模型AUC=0.886。
临床应用与阅片要点
- 在超声应变分析中,可关注游离壁(尤其侧壁与前游离壁)径向应变以提示肥胖相关心肌改变。
- 评估应激相关心肌受累时,可重点关注前壁中层纵向应变率与后间隔壁径向应变。
- 节段应变的区域分布模式有助于区分不同应激病因,而非仅依赖整体应变值。
局限性
- 研究为小鼠模型,结论外推至人类需谨慎。
- 样本量有限(每组n=24),机器学习模型未在独立队列验证。
- 未报告常规超声心动图参数(如射血分数)的变化。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
英文原文(摘要)· 点击展开
We used segmental strain analysis to evaluate whether intrinsic (diet-induced obesity [DIO]) and extrinsic (unpredictable chronic mild stress [UCMS]) stressors can alter deformational patterns of the left ventricle. Six-week-old male C57BL/6J mice were randomized into the lean or obese group (n = 24/group). Mice underwent 12 wk of DIO with a high-fat diet (HFD). At 18 wk, lean and obese mice were further randomized into UCMS and non-UCMS groups (UCMS, 7 h/d, 5 d/wk, for 8 wk). Echocardiography was performed at baseline (6 wk), post-HFD (18 wk) and post-UCMS (26 wk). Machine learning was applied to the DIO and UCMS groups. There was robust predictive accuracy (area under the receiver operating characteristic curve [AUC] = 0.921) when comparing obese with lean mice, with radial strain changes in the lateral (-64%, p ≤ 0.001) and anterior free (-53%, p < 0.001) walls being most informative. The ability to predict mice that underwent UCMS, irrespective of diet, was assessed (AUC = 0.886), revealing longitudinal strain rate of the anterior midwall and radial strain of the posterior septal wall as the top features. The wall segments indicate a predilection for changes in deformation patterns to the free wall (DIO) and septal wall (UCMS), indicating disease-specific alterations to the myocardium.
引用
- Kunovac, Hathaway, Burrage, et al. Left Ventricular Segmental Strain Identifies Unique Myocardial Deformation Patterns After Intrinsic and Extrinsic Stressors in Mice. Ultrasound Med Biol. 2022. PMID: 35933241. doi:10.1016/j.ultrasmedbio.2022.06.004
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。