Electrocardiography for the detection of left ventricular hypertrophy in an elderly population with long-standing hypertension in primary care: a secondary analysis of the CHELLO cohort study

Automated Classification of Left Ventricular Hypertrophy on Cardiac MRI

Left ventricular hypertrophy is an independent predictor of coronary artery disease, stroke, and heart failure. Our aim was to detect LVH cardiac magnetic resonance (CMR) scans with automatic methods. We developed an ensemble model based on a three-dimensional version of ResNet. The input of the network included short-axis and long-axis images. We also introduced a standardization methodology to unify the input images for noise reduction. The output of the network is the decision whether the patient has hypertrophy or not. We included 428 patients (mean age: 49 ± 18 years, 262 males) with LVH (346 hypertrophic cardiomyopathy, 45 cardiac amyloidosis, 11 Anderson–Fabry disease, 16 endomyocardial fibrosis, 10 aortic stenosis). Our control group consisted of 234 healthy subjects (mean age: 35 ± 15 years; 126 males) without any known cardiovascular diseases. The developed machine-learning-based model achieved a 92% F1-score and 97% recall on the hold-out dataset, which is comparable to the medical experts. Experiments showed that the standardization method was able to significantly boost the performance of the algorithm. The algorithm could improve the diagnostic accuracy, and it could open a new door to AI applications in CMR.

Performance of a novel ECG criterion for improving detection of left ventricular hypertrophy: a cross-sectional study in a general Chinese population

OBJECTIVES

The sensitivity of ECG for detecting left ventricular hypertrophy (LVH) is low. The aim of this study was to explore a better ECG criterion for screening LVH in a large general Chinese population.

DESIGN

Case-control study.

SETTING

China Medical University in Shenyang, China.

PARTICIPANTS

All permanent residents in Dawa, Zhangwu and Liaoyang aged 35 years or older were invited. Participants with unqualified data, pacemaker rhythm, frequent premature ventricular beats, Wolff-Parkinson-White syndrome, complete bundle branch block, myocardial infarction or hypertrophic cardiomyopathy were excluded. A total of 10 360 subjects (4630 males) were recruited.

INTERVENTIONS

A novel ECG criterion (Northeast China Rural Cardiovascular Health Study (NCRCHS)) composed of different ratios of maximum R wave in lead V₅ or V₆ (R_V5/V6), S wave in lead V₃ (S_V3) and R wave in lead aVL (R_aVL) was proposed and validated using multiple linear regression. Receiver-operating characteristic curves were used to compare the NCRCHS criterion with traditional criteria for LVH detection.

RESULTS

An optimised model (15^*R_aVL+8^*S_V3+7^*R_V5/V6) was constructed (R² 0.192, p<0.001) with the cut-off values of 36.8 mV for males and 26.1 mV for females. The maximum area under the curve was obtained using the NCRCHS criterion (male 0.74, 95% CI 0.73 to 0.75; female 0.73, 95% CI 0.72 to 0.75), followed by Cornell voltage criterion, Sokolow-Lyon criterion, Peguero-Lo Presti criterion, multi-ethnic study of atherosclerosis (MESA)-specific criterion and Syst-Eur voltage criterion. Compared with the Cornell voltage criterion, the NCRCHS criterion had a significantly higher sensitivity for detecting LVH at the same level of specificity (p<0.05).

CONCLUSIONS

The NCRCHS criterion significantly improved sensitivity for LVH detection in a general Chinese population, with cut-off values of 36.8 and 26.1 mV for males and females, respectively. This criterion can detect LVH earlier and better and may prevent subsequent cardiovascular diseases.