L.G. Akulov, R.V. Litovkin, Yu.P. Mukha

It is given the real time notion in electrophysiology by examples of EKG and EEG signals. For EKG in particular, time delay must be less than one cardio cycle, i.e. less than 1 second. It is proposed methods for time series filtering based on fast discrete wavelet transform. After forward wavelet transform some of coefficients are zeroing. Next step is inverse wavelet transform after that step most of noise signals are vanished. Under review is low pass filters and band pass filters. Examples of filters parameters estimation are given for EKG and EEG signals. Filters computation is based on dynamic error functional minimization. Functional is power of differential (difference between ideal and reconstructed signal) signal to power of noisy signal ratio. For filters computed amplitude-frequency responses what comparing with filters based on finite impulse response. It is given equations for transform complexity estimation for wavelet based filters and filters based on finite impulse response. It is shown that