Current recommendations for CRT device implantation rely upon the assessment of the symptomatic status of patients, LVEF, and QRS width. Nevertheless, 30% to 40% of patients who receive CRT according to guidelines are nonresponders to treatment. Kalscheur et al. showed that the application of ML to CRT can provide better classification of patients than the simple assessment of LBBB morphology and QRS duration. Similarly, we were able to demonstrate that the application of the RF method to electrocardiographic data and the dynamic analysis of LV strain curves significantly improves the prediction of CRT response. The relevance of imaging-derived parameters in the selection of CRT candidates remains underestimated. In the present study, the application of supervised ML approaches was able to identify groups of features that are good predictors of CRT response.
In our study, SF, ApR, and IHD were the most important variables associated with CRT response. These results are in line with an increasing amount of data underscoring the relevance of the visual assessment of LV mechanical discoordination and ischemic cardiomyopathy in determining in CRT response.
The application of machine learning (ML) methods to a set of common clinical, electrocardiographic, and echocardiographic data emphasizes the importance of a multiparametric approach for both the identification of CRT response and the prediction of prognosis after CRT. Our results underscore the importance of right ventricular function on both CRT response and outcome and the pivotal role of the global assessment of heart function in patients undergoing CRT. Despite these interesting results, additional studies on a broader population will be necessary to fully validate and understand the clinical application of ML to CRT. Although there remains much to understand, as demonstrated in our study, it is likely that the application of ML-derived algorithms will allow the stratification of CRT candidates to guide patients toward specific therapeutic approaches and additional focused clinical trials.
Read more in:
Importance
of Systematic Right Ventricular Assessment in Cardiac Resynchronization Therapy
Candidates: A Machine Learning Approach - PubMed (nih.gov)
J Am Soc Echocardiogr. 2021 May;34(5):494-502
Elena Galli, Virginie Le Rolle, Otto A Smiseth, Jurgen
Duchenne, John M Aalen, Camilla K Larsen, Elif A
Sade, Arnaud Hubert, Smitha Anilkumar, Martin
Penicka , Cecilia Linde, Christophe Leclercq, Alfredo
Hernandez, Jens-Uwe Voigt, Erwan Donal
PMID: 33422667
DOI: 10.1016/j.echo.2020.12.025
Shared with permission from Elsevier under license number 5514750975673
