Reproduce “Real-time ultra-low power ECG anomaly detection using an event-driven neuromorphic processor” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Felix Christian Bauer, Dylan Richard Muir, Giacomo Indiveri
A system for real-time, ultra-low power ECG anomaly detection leverages an event-driven neuromorphic processor to process multi-channel ECG data locally. This approach achieved 91.3% sensitivity and 97.6% specificity in detecting cardiac anomalies, operating at less than 722.1 µW total power consumption.
Key insights
- Neuromorphic processors can effectively implement real-time, ultra-low power processing of complex biomedical signals like ECG, bridging the gap between SNN simulations and practical hardware deployment.
- The event-driven nature and sparse activity of SNNs on specialized hardware enable sub-milliwatt power consumption, making continuous "always-on" monitoring feasible for wearable devices.
- On-device preliminary diagnosis by the system reduces the need for extensive data storage and transmission, minimizing energy consumption and alleviating diagnostic burden on medical professionals.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.