Simulating 'Green AI' for Wearable Heart Monitors

Master Project   Ongoing

Project Overview

Current wearable health devices, such as smartwatches and portable biosensors, are increasingly expected to perform real-time health monitoring and intelligent diagnostics. However, many modern AI techniques remain too computationally expensive for low-power embedded platforms. This project investigates energy-efficient machine learning approaches for biomedical signal analysis using Reservoir Computing techniques applied to ECG-based arrhythmia classification.

The project focuses on developing lightweight AI models capable of analysing noisy physiological signals while maintaining low computational complexity and power consumption. Using real-world biomedical datasets, the work explores how unconventional machine learning architectures can provide accurate cardiac anomaly detection suitable for future wearable and edge-AI healthcare systems.

Graduate Researcher

  • Abdul Hanan Ahmad

Supervisory Team