Signal Quality Index (SQI) Algorithms for PPG
Signal Quality Index (SQI) algorithms automatically grade PPG signal reliability on a 0–1 scale using template matching, spectral analysis, morphological validation, and statistical tests, enabling downstream algorithms to process only high-quality data and suppress false measurements.
SQI computation combines multiple quality dimensions. Template-based SQI computes the Pearson correlation between each detected beat and a running average template from the same recording; r < 0.9 flags potential artifacts. Spectral SQI measures the ratio of power in the expected cardiac band to total signal power; cardiac band ratios below 0.4 indicate excessive noise. Morphological SQI validates physiological plausibility of pulse shape (positive systolic peak, reasonable width, amplitude within expected range).
Ensemble SQI frameworks combining 3–5 individual metrics via weighted averaging or decision trees outperform single-metric approaches. The PhysioNet 2014 Challenge established that ensemble SQI achieves AUC > 0.95 for artifact detection. Optimal ensemble weights are data-dependent: template correlation receives highest weight for motion artifacts, spectral SQI for electrical interference, and morphological SQI for sensor detachment events.
For real-time wearable implementation, lightweight SQI computations run continuously alongside PPG sampling, providing beat-level or epoch-level quality annotations. Beats with SQI < 0.8 are excluded from HR, HRV, and SpO2 calculations. This quality gating reduces false alarm rates in clinical monitoring by 60–80% and improves consumer wearable HR accuracy by 20–40% by preventing erroneous measurements during motion episodes.
Frequently Asked Questions
What SQI threshold should be used for clinical applications?
SQI > 0.9 is recommended for clinical HRV analysis and SpO2 monitoring. SQI > 0.8 is acceptable for heart rate estimation. SQI > 0.7 may be used for trend monitoring where individual measurement precision is less critical.
Can SQI predict when PPG measurement will fail?
Trending SQI over time can predict impending signal loss: declining SQI from 0.95 to 0.80 over 30 seconds typically precedes complete signal loss from sensor displacement. Predictive SQI alerts can prompt sensor repositioning before data loss.
Does SQI need to be device-specific?
SQI algorithms are partially device-specific because PPG waveform characteristics (amplitude, shape, noise profile) vary across sensor designs. Template-based SQI is the most transferable method because templates are derived from the same device's clean signals.