Cardiac pre-ejection period detection

Claim 3

The method of claim 1 wherein an allowable time to determine the time of best AR match is confined to a time window that begins with the peak of the R wave, which gives rise to ejection during said on-going cardiac beat, and ends with the end of a main lobe of the RMS-PEP waveform corresponding to said beat.

Claim 4

The method of claim 1 further comprising the steps of: (a) preconditioning the accelerometer signal; (b) processing a first segment of the preconditioned accelerometer signal of (a) by algorithms selected from the group consisting of Levinson and Yule-Walker from autocorrelation lags of the accelerometer signal derived during the time of a main lobe of the RMS-PEP waveform; (c) subjecting the output of the processed signal of (b) to an inverse filter; (d) extracting the output of the inverse filter as an absolute value; (e) processing a second segment of the preconditioned accelerometer signal of (a) to produce a running RMS estimate of the accelerometer signal; (f) scaling the absolute value of the output of the inverse filter by the running RMS estimate of the accelerometer signal during the main lobe of the RMS-PEP waveform to produce a scaled signal; (g) using a peak detector to detect the peak absolute value of the scaled signal in (f) to determine the time of minimum prediction error, which is the time of best AR match; and (h) subtracting the time of the R wave peak from the time of best AR match to obtain PEP.

Claim 5

The method of claim 3 wherein the step of subjecting the output of the processed signal of (b) to an inverse filter includes subjecting same to an FIR lattice filter.

Claim 6

The method of claim 3 wherein an allowable time to determine the time of best AR match is confined to a time window that begins with the peak of the R wave, which gives rise to ejection during said on-going cardiac beat, and ends with the end of a main lobe of the RMS-PEP waveform corresponding to said beat.

Claim 7

The method of claim 3 wherein the step of preconditioning the accelerometer signal further comprises the steps of (a) subjecting the accelerometer signal to a low pass filter; (b) extracting the first derrivative of the output of the low pass filter; and (c) digitizing the signal using an A-to-D converter.

Claim 8

In a cardiac pre-ejection period detector for use in controlling a rate adaptive pacer based on the pre-ejection period (PEP) said pacer having an ECG signal detector including an R wave peak detector means and an associated accelerometer transducer having an output signal: (a) means for preconditioning the accelerometer output signal; (b) means for processing a first segment of the preconditioned accelerometer output signal of (a) to produce a running RMS-PEP waveform of the accelerometer output signal; (c) means for processing a second segment of the preconditioned accelerometer output signal of (a) using algorithms selected from the group consisting of Levinson and Yule-Walker from autocorrelation lags of the accelerometer signal derived during the time of a main lobe of the RMS-PEP signal to produce an auto-regressive (AR) signal output; (d) inverse filter means for processing said AR signal output and producing an inverse filter output signal; (e) means for producing an absolute value signal from the inverse filter output signal; (f) multiplying means for scaling the absolute value of the output of the inverse filter by the running RMS-PEP waveform of the accelerometer signal during the main lobe of the RMS-PEP waveform; (g) detector means to detect the peak absolute value of the scaled inverse filter output signal to determine the time of minimum prediction error which is the best AR match; and (h) means for subtracting the time of the peak of the R wave from the time of the best AR match to obtain PEP.

Claim 9

In a cardiac pre-ejection period detector for use in controlling a rate adaptive pacer based on the pre-ejection period (PEP) said pacer having an ECG signal detector including an R wave peak detector means and an associated accelerometer transducer having an output signal: (a) means for preconditioning the accelerometer output signal; (b) means for processing a first segment of the preconditioned accelerometer output signal of (a) to produce a running RMS-PEP waveform of the accelerometer output signal; (c) means for processing a second segment of the preconditioned accelerometer output signal of (a) using algorithms selected from the group consisting of Levinson and Yule-Walker from autocorrelation lags of the accelerometer signal derived during the time of a main lobe of the RMS-PEP signal to produce an auto-regressive (AR) signal output; (d) inverse filter means for processing said AR signal output and producing an inverse filter output signal; (e) means for producing an absolute value signal from the inverse filter output signal; (f) multiplying means for scaling the absolute value of the output of the inverse filter by the running RMS-PEP waveform of the accelerometer signal during the main lobe of the RMS-PEP waveform; (g) detector means to detect the peak absolute value of the scaled inverse filter output signal to determine the time of minimum prediction error which is the best AR match; and (h) means for subtracting the time of the peak of the R wave from the time of the best AR match to obtain PEP.

Claim 10

The apparatus of claim 7 wherein said inverse filter is an FIR lattice filter.