Decomposition and modeling of signal shapes of single point cardiac monitoring

Abstract

We introduce a novel method for electronic recording of cardiac signals from a single point at the skin in contrast to classical differential electrocardiography (ECG). Ultralownoise transistor devices with an adaptable, auto-stabilizing transimpedance amplifier are able to measure tiny skin potential modulations from a single contact electrode located at an individual’s wrist (single-point cardiography-SPC). Although SPC signals were highly prone to interspersed noise, they contained periodic patterns. In an electromagnetically shielded setting, we could clearly extract breathing and cardiac rhythms from the acquired SPC signals. As the reference, we measured ECG in parallel. Several signal-processing techniques like smoothening, correlation, decomposition and signal extraction showed that SPC signals contain breathing and periodic heart potential variations, which are time-correlated with ECG. In the future, we intent to use this novel technique to measure heart signals from patients in different health conditions.

Keywords

• hodrick-prescott decomposition
• arima
• stl
• signal decomposition
• algan/gan hemt