Wearable System for Cardiac Diagnosis and Monitoring: Clustering Analysis and Usability Assessment Using Fractal Geometry

Abstract

Heart rate abnormalities, including tachycardia and bradycardia, are among the leading global causes of morbidity, necessitating continuous and accurate monitoring for early detection and intervention. This study introduces an innovative smartwatch-based cardiac monitoring system that integrates fractal geometry analysis, dynamical systems modeling, and clustering techniques to enhance diagnostic precision. Unlike conventional smartwatch-based monitoring systems, this approach employs advanced mathematical modeling to identify nonlinear patterns in heart rate dynamics, enabling more precise differentiation between normal and pathological conditions. The system was developed using the CRISP-DM methodology, ensuring a structured and data-driven implementation. A mobile application, “My Cardio,” was designed for Android-based smartwatches, enabling the collection of real-time heart rate data and cloud storage for subsequent fractal-based processing. Additionally, a clustering analysis was performed on data from patients with and without cardiac history, identifying three distinct patient groups based on heart rate characteristics. Cluster 0 included individuals with lower, stable heart rates; Cluster 1 represented intermediate variations; and Cluster 2 comprised patients with significantly elevated heart rates associated with higher clinical risk. The findings were statistically validated and visualized, demonstrating that integrating clustering techniques with fractal geometry enhances the detection of clinically relevant cardiac patterns. Furthermore, a usability assessment using the System Usability Scale (SUS) yielded a score of 80.3 or higher, confirming high user acceptance and feasibility for widespread adoption. This study differentiates itself from existing approaches by combining wearable technology with advanced computational techniques to enhance cardiac diagnosis and monitoring. The results underscore the potential of smartwatch-based systems as a noninvasive, intelligent alternative for continuous cardiovascular assessment, paving the way for future applications in digital cardiology and telemedicine.

Keywords

• Clustering
• fractal geometry
• cardiac monitoring
• smartwatch
• usability