Localization of Ventricular Activation Origin from the 12-Lead ECG: A Comparison of Linear Regression with Non-Linear Methods of Machine Learning

Zhou, Shijie; AbdelWahab, Amir; Sapp, John L.; Warren, James W.; Horáček, B. Milan

doi:10.1007/s10439-018-02168-y

Cited by 24 publications

(25 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…No one ML algorithm is the most accurate in all cases, and thus comparisons of ML algorithms in different fields of research, and on different datasets, may yield different results 37 . In this study, a comparison of the regression models showed that other linear models performed similarly on the test set to the MLR model, a frequently used and simple statistical model that makes it easy to interpret the variables 38 . A possible explanation for this is that these linear models had the same optimal feature set (see Supplementary Tables S12-S15), and no multi-collinear relationship was obtained between the features (see Fig.…”

Section: Discussionmentioning

confidence: 75%

See 1 more Smart Citation

A machine learning approach to personalized dose adjustment of lamotrigine using noninvasive clinical parameters

Zhu

Huang

et al. 2021

Sci Rep

View full text Add to dashboard Cite

The pharmacokinetic variability of lamotrigine (LTG) plays a significant role in its dosing requirements. Our goal here was to use noninvasive clinical parameters to predict the dose-adjusted concentrations (C/D ratio) of LTG based on machine learning (ML) algorithms. A total of 1141 therapeutic drug-monitoring measurements were used, 80% of which were randomly selected as the "derivation cohort" to develop the prediction algorithm, and the remaining 20% constituted the "validation cohort" to test the finally selected model. Fifteen ML models were optimized and evaluated by tenfold cross-validation on the "derivation cohort,” and were filtered by the mean absolute error (MAE). On the whole, the nonlinear models outperformed the linear models. The extra-trees’ regression algorithm delivered good performance, and was chosen to establish the predictive model. The important features were then analyzed and parameters of the model adjusted to develop the best prediction model, which accurately described the C/D ratio of LTG, especially in the intermediate-to-high range (≥ 22.1 μg mL−1 g−1 day), as illustrated by a minimal bias (mean relative error (%) = + 3%), good precision (MAE = 8.7 μg mL−1 g−1 day), and a high percentage of predictions within ± 20% of the empirical values (60.47%). This is the first study, to the best of our knowledge, to use ML algorithms to predict the C/D ratio of LTG. The results here can help clinicians adjust doses of LTG administered to patients to minimize adverse reactions.

show abstract

Section: Discussionmentioning

confidence: 75%

“…Moreover, many factors affect the outputs, particularly in nonlinear, dynamic disease states. Therefore, the dataset may inevitably contain noise that can affect linear regression models more than nonlinear regression models 38 .…”

Section: Discussionmentioning

confidence: 99%

A machine learning approach to personalized dose adjustment of lamotrigine using noninvasive clinical parameters

Zhu

Huang

et al. 2021

Sci Rep

View full text Add to dashboard Cite

show abstract

“…In addition, the accuracy of PAVEL could contribute an error to VT exit site localization. PAVEL has been shown to localize the 3D coordinates of a pacing site on the LV endocardium with an accuracy of 12.5 mm 16 . This error was for localizing a pacing site and not a VT exit site; in the latter case, we expect a higher margin of error.…”

Section: Discussionmentioning

confidence: 92%

“…The PAVEL system combines information from an acquired eight‐lead ECGs (leads I, II, V1‐V6) of an induced VT and population‐derived regression coefficients to identify the site of earlies LV activation onto the generic LV endocardial surface. A total of 1012 pacing sites, each of which were manually registered onto one of 238 triangles whose coordinate centers of the generic LV endocardial surface and their corresponding eight‐lead ECGs, constituted a training set for computing population‐derived regression coefficients 8,16 . For the acquired ECG of the induced VT, the PAVEL system calculates the 120‐ms ∫QRS of eight leads (Figure 1, panel E), and uses the population‐derived regression coefficients determined in a multiple regression model to predict the VT exit site coordinates ( x, y, z ) from the eight‐lead 120‐ms ∫QRS of the VT.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Feasibility study shows concordance between image‐based virtual‐heart ablation targets and predicted ECG‐based arrhythmia exit‐sites

Zhou

Sung

Prakosa

et al. 2021

Pacing Clinical Electrophis

Self Cite

View full text Add to dashboard Cite

Introduction We recently developed two noninvasive methodologies to help guide VT ablation: population‐derived automated VT exit localization (PAVEL) and virtual‐heart arrhythmia ablation targeting (VAAT). We hypothesized that while very different in their nature, limitations, and type of ablation targets (substrate‐based vs. clinical VT), the image‐based VAAT and the ECG‐based PAVEL technologies would be spatially concordant in their predictions. Objective The objective is to test this hypothesis in ischemic cardiomyopathy patients in a retrospective feasibility study. Methods Four post‐infarct patients who underwent LV VT ablation and had pre‐procedural LGE‐CMRs were enrolled. Virtual hearts with patient‐specific scar and border zone identified potential VTs and ablation targets. Patient‐specific PAVEL based on a population‐derived statistical method localized VT exit sites onto a patient‐specific 238‐triangle LV endocardial surface. Results Ten induced VTs were analyzed and 9‐exit sites were localized by PAVEL onto the patient‐specific LV endocardial surface. All nine predicted VT exit sites were in the scar border zone defined by voltage mapping and spatially correlated with successful clinical lesions. There were 2.3 ± 1.9 VTs per patient in the models. All five VAAT lesions fell within regions ablated clinically. VAAT targets correlated well with 6 PAVEL‐predicted VT exit sites. The distance between the center of the predicted VT‐exit‐site triangle and nearest corresponding VAAT ablation lesion was 10.7 ± 7.3 mm. Conclusions VAAT targets are concordant with the patient‐specific PAVEL‐predicted VT exit sites. These findings support investigation into combining these two complementary technologies as a noninvasive, clinical tool for targeting clinically induced VTs and regions likely to harbor potential VTs.

show abstract

A “two-step classification” machine learning method for non-invasive localization of premature ventricular contraction origins based on 12-lead ECG

Wang

Feng

Zhong

et al. 2023

J Interv Card Electrophysiol

View full text Add to dashboard Cite

Localization of Ventricular Activation Origin from the 12-Lead ECG: A Comparison of Linear Regression with Non-Linear Methods of Machine Learning

Cited by 24 publications

References 26 publications

A machine learning approach to personalized dose adjustment of lamotrigine using noninvasive clinical parameters

A machine learning approach to personalized dose adjustment of lamotrigine using noninvasive clinical parameters

Feasibility study shows concordance between image‐based virtual‐heart ablation targets and predicted ECG‐based arrhythmia exit‐sites

A “two-step classification” machine learning method for non-invasive localization of premature ventricular contraction origins based on 12-lead ECG

Contact Info

Product

Resources

About