Design and Performance Evaluation of an AI-Assisted Low-Cost 3-Lead ECG System Using a Modified Chest-Based Einthoven Configuration

Background: Electrocardiography (ECG) is a gold standard for cardiac assessment. While conventional 12- lead ECG systems provide comprehensive spatial information, they are costly, complex, and unsuitable for continuous or wearable monitoring. Simplified 3-lead ECG systems offer a practical alternative for preliminary screening and long-term observation. Recent advances in low-cost hardware and artificial intelligence (AI) enable intelligent ECG systems suitable for portable applications. Aim: This study aims to design, implement, and evaluate a low-cost, AI-assisted 3-lead ECG system using Arduino and the AD8232 sensor, employing a modified chest-based Einthoven configuration, and to assess its ability to differentiate between normal and abnormal ECG patterns. Methodology: A prototype ECG system was developed using an Arduino Uno and AD8232 analog front-end. Three electrodes were placed on the chest to approximate limb-lead vectors, forming a modified Einthoven triangle. ECG signals were acquired at 360 Hz and processed using Python-based signal processing techniques, including adaptive filtering and automated feature extraction. System validity was assessed using a modified form of Einthoven’s law. A comparative analysis was performed between normal and abnormal ECG groups using quantitative parameters and statistical testing. Results: The system successfully acquired stable ECG signals across three leads, with Lead II demonstrating the highest signal-to-noise ratio (SNR). Validation of the modified Einthoven relationship showed a low error of 1.13%, indicating accurate lead derivation despite chestbased placement. Comparative analysis revealed significant differences between normal and abnormal ECG patterns in heart rate, RR interval variability, and QRS duration (t = 4.21, p < 0.01). Signal processing improved SNR by more than 100%. Conclusion: The proposed system demonstrates that a low-cost, chest-based 3-lead ECG configuration combined with AI-assisted analysis can provide reliable cardiac monitoring. While not a replacement for 12-lead ECG systems, it is a promising solution for wearable, educational, and remote healthcare applications.