ADAPTIVE ALGORITHMS FOR REAL-TIME DETECTION OF CARDIAC RHYTHM DISORDERS

Abstract

Early detection of cardiac rhythm disorders is an important task for telemedicine, remote monitoring, and intelligent medical devices. In particular, real-time processing of cardiac signals, automatic detection of dangerous rhythm changes, and rapid notification in emergency situations significantly increase the practical value of the system. Recent studies show that there are three main challenges in ECG-based automatic arrhythmia detection: distortion of the signal by dynamic noise, inter-patient physiological variability, and resource constraints in edge devices [1]–[4]. This paper proposes an adaptive algorithm for a Holter-based monitoring system operating on the Raspberry Pi platform and automatically sending emergency alerts via GSM technology. The proposed approach includes signal preprocessing, R-peak detection, RR-interval calculation, rhythm evaluation based on adaptive thresholds, two-stage confirmation, and alert transmission for confirmed abnormal events. The aim of the approach is to reduce false alarms, account for individual physiological differences, and ensure resource-efficient real-time operation.