Pulse Wave Peak Detection via Amplitude Ratio Filtering

The patent applies preliminary action by performing moving average processing on the pulse wave data before peak detection. This preprocessing step smooths the waveform and removes notch noise components, creating cleaner data for subsequent peak detection. The moving average calculation is performed over a predetermined number of data points to eliminate high-frequency noise while preserving the underlying pulse wave characteristics.

The patent segments the pulse wave data into discrete data points along the time axis, allowing individual processing of each point through the moving average algorithm. By dividing the continuous waveform into separable data elements, the method can apply statistical processing to distinguish true peaks from noise components, improving detection accuracy without requiring complex analog filtering.

The patent uses pulse wave data as an intermediary to indirectly obtain RR interval information. Instead of directly measuring ECG signals which require electrode contact, the method detects pulse waves at peripheral sites (such as finger or wrist) which are synchronized with cardiac activity. This intermediary approach maintains measurement precision while significantly improving subject comfort and ease of operation.

The patent replaces the mechanical/electrical contact system of ECG electrodes with a non-contact or minimal-contact pulse wave detection system. By substituting the direct electrical measurement approach with optical or pressure-based pulse wave sensing, the method eliminates the need for multiple electrodes and adhesive contacts, thereby improving ease of operation while preserving the ability to extract cardiac timing information.