High-Fidelity ECG Analysis // Consumer Sensors
build_target: android_8.0+ | status: experimental
Potential Class 2 Medical Device. Not cleared for clinical use.
> Contact h10s@andrewhowden.com for access keys.
./abstract.txt
The h10s project investigates the efficacy of using consumer-grade ECG hardware (Polar H10, Movesense MD) for high-fidelity cardiac rhythm analysis. Bypassing standard averaging algorithms, we access raw sensor data at 512Hz. Real-time beat morphology clustering in user-space.
Methodology
Direct signal processing pipeline to ensure clinical-grade precision. No black-box averaging.
Signal Acquisition
Interfacing via BLE L2CAP/ATT. Raw ECG streams at 130Hz - 512Hz. Preserving QRS complex morphology.
On-Device Processing
Zero cloud dependency.
- Pan-Tompkins Algorithm (QRS detection)
- Unsupervised Beat Clustering (k-means adaptation)
- Output Format: EDF+ (Data Interoperability)
Captured_Artifacts
