Wearable HRV Variability Weighting for Reliable Sleep Metrics

The patent applies preliminary action by calculating the variability metric of HRV data before using it to determine physiological parameters. The system first computes the variability metric (using interquartile range, coefficient of variation, or standard deviation) as a preliminary step, then uses this metric to weight or filter HRV values before final parameter calculation. This preliminary assessment of variability ensures that outlier-prone data is identified and handled appropriately before affecting the final health insights, thereby improving reliability without adding excessive complexity to the overall system.

The patent applies parameter changes by introducing a variability metric as an additional parameter that characterizes the HRV data quality. Instead of directly using raw HRV values, the system transforms the data by computing the variability metric (interquartile range, coefficient of variation, or standard deviation) and then uses this transformed parameter to weight or select appropriate HRV values. This parameter transformation approach improves measurement precision while keeping the computational complexity manageable, as the additional calculations are straightforward statistical operations that can be efficiently implemented in wearable devices.