Scaling factor and jitter characterization: an unified approach using mean and variance

In signal averaging of recurrent signals, like ECGs, EPs and EMGs, the averaged signal can be represented by a convolution equation involving the ideal mean signal and the jitter effect. When the jitter variation is replaced by a random scaling factor, the averaged signal is not issued from the same processing. Here, the authors precisely determine which equation leads to the averaged signal when a scaling factor is present instead of jitter. In both cases, the mean of the realizations is estimated for each time. The authors show the effects of the scaling factor and the jitter in the estimation of the realized variances for each time sample. The characterization of these effects will lead to estimations of both scaling and delay variances.