Apparatus and method for estimating missing doppler signals and spectra

Abstract

The invention provides a method and system for estimating missing Doppler signals and spectra in an ultrasound imaging system: (1) parameters representative of a Doppler data model are matched to actual Doppler data that is received; (2) the parameters are interpolated or extrapolated (possibly after normalization, smoothing, or other manipulation) for regions where actual Doppler data is missing ('Doppler gaps'); and (3) a set of synthesized Doppler spectra are constructed, using the model and the interpolated or extrapolated parameters, for those Doppler gaps. The Doppler data model is an autoregressive model, or another model in which the Doppler data may be approximated by supplying a source signal to an autoregressive filter. The parameters of the model comprise a set of autoregressive filter coefficients and a .sigma..sup.2 (variance) value. The reflection coefficients of these autoregressive filter coefficients are determined. The reflection coefficients and .sigma..sup.2 value for a Doppler gap are interpolated from reflection coefficients and .sigma..sup.2 values from at least one block of Doppler data before the Doppler gap and at least one block of Doppler data after the Doppler gap. In a preferred embodiment, a windowing function may be used when combining the actual Doppler data with the synthesized Doppler data. A set of synthesized Doppler spectra may also be determined in response to the Doppler data near the Doppler gap. A white noise source having the interpolated .sigma..sup.2 value may be generated and filtered using an autoregressive filter having the interpolated reflection coefficients, to produce a set of synthesized Doppler data.