Table 4.

Summary table for advantages and disadvantages of data imputation methods.

Method	Advantages	Disadvantages
Linear	- Assumes less than the other methods - Simple and efficient for good quality signals	- Less effective for signals with lots of missing data - Loss of time dependency
Pchip	- Preserves the linear trend and the slightly non linear contributions in the RR time-series [32]	- Less effective for signals with lots of missing data - Loss of time dependency
Spline	- Can capture abrupt variations when data quality is good	- Introduces outliers due to oscillation of the interpolation function [9] - Less effective for signals with lots of missing data - Loss of time dependency
DVC	- Adaptive to data distribution and variability - No ectopic values in the processed signal - Preserves signal’s time dependency - Effective for low quality signals	- Computationally expensive - Algorithm could be optimised