Design and Implementation of a Cloud-Integrated Desktop ECG System Using a Multi-Layer Perceptron for Arrhythmia Classification

Cardiovascular diseases (CVDs) remain the foremost cause of mortality globally, necessitating the development of advanced tools for early and accurate cardiac diagnosis. This paper presents the comprehensive design, implementation, and evaluation of a desktop-based Electrocardiogram (ECG) monitoring system. The system architecture integrates a powerful Multi-Layer Perceptron (MLP) deep learning model designed to automatically identify and classify critical heart rhythm abnormalities, including bradycardia, tachycardia, and other forms of arrhythmia. A cornerstone of this system is its seamless and secure integration with a Supabase cloud backend, which facilitates centralized data storage, real-time synchronization, and secure, role-based access for various healthcare professionals, rigorously enforced through PostgreSQL’s Row Level Security (RLS). The MLP model was trained and validated on a diverse and extensive collection of data from the MITBIH Arrhythmia, PTB Diagnostic ECG, and Kaggle databases. Empirical evaluation results demonstrate high model performance, with classification accuracies reaching 92% for both bradycardia and tachycardia, and 89% for general arrhythmia detection. Functional and performance testing further validate the system’s operational reliability, showing an average cloud data synchronization time of approximately 4 seconds and robust, though partially incomplete, RLS policy enforcement. This work contributes a scalable, accurate, and secure solution for advanced cardiac monitoring in desktop environments, effectively bridging the gap between clinical-grade analysis and accessible, userfriendly technology

Keywords: Electrocardiogram, MLP, deep learning, arrhythmia, bradycardia, tachycardia, HRV, desktop health application, cloud computing, RLS, Supabase, Flutter

Authors who publish in this journal agree to the following rules:

1. Authors retain copyright and give the journal the right of first publication, with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
2. Authors may enter separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., posting it to an institutional repository or publishing it in a book), with attribution to the journal's initial publication.
3. Authors are permitted and recommended to post their work online (such as in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges as well as earlier and greater citation of published work.