Sudden cardiac death multiparametric classification system for Chagas heart disease's patients based on clinical data and 24-hours ECG monitoring

Abstract

<abstract><p>About 6.5 million people are infected with Chagas disease (CD) globally, and WHO estimates that $ > million people worldwide suffer from ChHD. Sudden cardiac death (SCD) represents one of the leading causes of death worldwide and affects approximately 65% of ChHD patients at a rate of 24 per 1000 patient-years, much greater than the SCD rate in the general population. Its occurrence in the specific context of ChHD needs to be better exploited. This paper provides the first evidence supporting the use of machine learning (ML) methods within non-invasive tests: patients' clinical data and cardiac restitution metrics (CRM) features extracted from ECG-Holter recordings as an adjunct in the SCD risk assessment in ChHD. The feature selection (FS) flows evaluated 5 different groups of attributes formed from patients' clinical and physiological data to identify relevant attributes among 57 features reported by 315 patients at HUCFF-UFRJ. The FS flow with FS techniques (variance, ANOVA, and recursive feature elimination) and Naive Bayes (NB) model achieved the best classification performance with 90.63% recall (sensitivity) and 80.55% AUC. The initial feature set is reduced to a subset of 13 features (4 Classification; 1 Treatment; 1 CRM; and 7 Heart Tests). The proposed method represents an intelligent diagnostic support system that predicts the high risk of SCD in ChHD patients and highlights the clinical and CRM data that most strongly impact the final outcome.</p></abstract>

Risk stratifier for sudden cardiac death beyond the left ventricular ejection fraction in Chagas cardiomyopathy

“…The analysis of the ECG signals data was based on the feature extraction and wave delineation system developed by Madeiro and colleagues 12,13 . The method was validated for reproducibility in two studies 6,7 and has been used in other studies 8 …”

“…In a previous study by our group, which has been detailed elsewhere 8 applying supervised machine learning to 315 patients (as output SCD and input 57 variables), with 80% for the training base and 20% for the validation, the Balanced Random Forest (BRF) classifier showed the best performance (with an accuracy of 82%, an AUC curve of 82%, a precision of 61%, an F1‐Score of 70%, and sensitivity of 82.5%) as a model for prediction of SCD (see complementary material). In the present study, only patients included in the previous study were considered.…”

“…12,13 The method was validated for reproducibility in two studies 6,7 and has been used in other studies. 8 The feature extraction steps were applied to ECG data between the 4 AM to 8 AM period (greater sympathetic effect) 14 for each patient.…”

“…In a previous study by our group, which has been detailed elsewhere 8 applying supervised machine learning to 315 patients S1).…”

“…We believe that a better understanding of the electrophysiological heterogeneity involved in SCD could better identify events treatable with ICD and, therefore, influence clinical decisions in this regard in CC. In a previous study, 8 we showed the relevance of the cardiac restitution metrics extracted from 4-h ECG segments as an adjuvant in assessing the risk of SCD in CC.…”