An ECG-Based Model for Left Ventricular Hypertrophy Detection: A Machine Learning Approach.

Area of Science:

• Cardiology
• Biomedical Engineering
• Artificial Intelligence in Medicine

Background:

• Left ventricular hypertrophy (LVH) is common, but current electrocardiography (ECG) diagnostic criteria have low sensitivity.
• There is a need for more accurate methods to detect LVH using ECG data.

Purpose of the Study:

• To develop and evaluate an automated method for LVH detection using ECG-extracted features and machine learning.
• To compare the performance of machine learning models against conventional clinical ECG criteria for LVH diagnosis.

Main Methods:

• ECG data from PTB-XL and Georgia databases were used, with features automatically extracted.
• Machine learning models (logistic regression, random forest, SVM) were trained on selected ECG features.
• Model performance was evaluated on a separate dataset and compared with standard clinical LVH-ECG criteria.

Main Results:

• Machine learning models, particularly Random Forest and Support Vector Machine, achieved over 90% accuracy.
• The proposed models demonstrated significantly improved sensitivity (above 86%) compared to clinical criteria (maximum 38%).

Conclusions:

• Automated LVH detection using machine learning and ECG features offers superior diagnostic performance over conventional methods.
• This advanced approach has the potential to enhance clinical practice for LVH diagnosis.