School of Biomedical Engineering, Dalhousie University, Dentistry Building, 5981 University Avenue, PO BOX 15000, Halifax, NS, B3H 4R2, Canada.
Department of Medicine, Dalhousie University, Halifax, NS, Canada.
Ann Biomed Eng. 2019 Feb;47(2):403-412. doi: 10.1007/s10439-018-02168-y. Epub 2018 Nov 21.
We have previously developed an automated localization method based on multiple linear regression (MLR) model to estimate the activation origin on a generic left-ventricular (LV) endocardial surface in real time from the 12-lead ECG. The present study sought to investigate whether machine learning-namely, random-forest regression (RFR) and support-vector regression (SVR)-can improve the localization accuracy compared to MLR. For 38 patients the 12-lead ECG was acquired during LV endocardial pacing at 1012 sites with known coordinates exported from an electroanatomic mapping system; each pacing site was then registered to a generic LV endocardial surface subdivided into 16 segments tessellated into 238 triangles. ECGs were reduced to one variable per lead, consisting of 120-ms time integral of the QRS. To compare three regression models, the entire dataset ([Formula: see text]) was partitioned at random into a design set with 80% and a test set with the remaining 20% of the entire set, and the localization error-measured as geodesic distance on the generic LV surface-was assessed. Bootstrap method with replacement, using 1000 resampling trials, estimated each model's error distribution for the left-out sample ([Formula: see text]). In the design set ([Formula: see text]), the mean accuracy was 8.8, 12.1, and 12.9 mm, respectively for SVR, RVR and MLR. In the test set ([Formula: see text]), the mean value of the localization error in the SVR model was consistently lower than the other two models, both in comparison with the MLR (11.4 vs. 12.5 mm), and with the RFR (11.4 vs. 12.0 mm); the RFR model was also better than the MLR model for estimating localization accuracy (12.0 vs. 12.5 mm). The bootstrap method with 1,000 trials confirmed that the SVR and RFR models had significantly higher predictive accurate than the MLR in the bootstrap assessment with the left-out sample (SVR vs. MLR ([Formula: see text]), RFR vs. MLR ([Formula: see text])). The performance comparison of regression models showed that a modest improvement in localization accuracy can be achieved by SVR and RFR models, in comparison with MLR. The "population" coefficients generated by the optimized SVR model from our dataset are superior to the previously-derived "population" coefficients generated by the MLR model and can supersede them to improve the localization of ventricular activation on the generic LV endocardial surface.
我们之前开发了一种基于多元线性回归（MLR）模型的自动定位方法，可从 12 导联心电图实时估算通用左心室（LV）心内膜表面上的激活起源。本研究旨在探讨机器学习（即随机森林回归（RFR）和支持向量回归（SVR））是否可以比 MLR 提高定位准确性。对于 38 名患者，在 LV 心内膜起搏期间使用从电解剖映射系统导出的已知坐标获取 12 导联心电图，共 1012 个起搏部位；然后将每个起搏部位注册到通用 LV 心内膜表面上，该表面被细分为 16 个分段，这些分段被细分为 238 个三角形。心电图被简化为每个导联一个变量，由 QRS 的 120ms 时间积分组成。为了比较三种回归模型，整个数据集（[公式：见文本]）随机分为设计集（80%）和测试集（20%），并评估定位误差（在通用 LV 表面上的测地线距离）。使用替换的 1000 次重采样试验的引导方法估计了每个模型对遗漏样本的误差分布（[公式：见文本]）。在设计集中（[公式：见文本]），SVR、RVR 和 MLR 的平均准确性分别为 8.8、12.1 和 12.9mm。在测试集中（[公式：见文本]），SVR 模型的定位误差平均值始终低于其他两种模型，与 MLR（11.4 对 12.5mm）和 RFR（11.4 对 12.0mm）相比；RFR 模型在估计定位准确性方面也优于 MLR 模型（12.0 对 12.5mm）。具有 1000 次试验的引导方法证实，SVR 和 RFR 模型在引导评估中比 MLR 具有更高的预测准确性（SVR 对 MLR（[公式：见文本]），RFR 对 MLR（[公式：见文本]））。回归模型的性能比较表明，与 MLR 相比，SVR 和 RFR 模型可以适度提高定位准确性。从我们的数据集中优化的 SVR 模型生成的“总体”系数优于之前从 MLR 模型生成的“总体”系数，并且可以取代它们来改善通用 LV 心内膜表面上的心室激活定位。
