Variable mode averager
First Claim
1. A variable mode averager comprising:
- a buffer that stores a plurality of weighted input values;
a mode input that specifies a time value relative to said input values; and
a processor coupled to said buffer, said processor configured to provide an estimate of said input values, said estimate corresponding to said time value.
6 Assignments
0 Petitions
Accused Products
Abstract
A signal processor generates an estimate of a block of weighted input values. An adjustable mode parameter determines a time point relative to the input values at which the estimate is computed. By varying the mode parameter, the output characteristics of the processor are variable between that of a smoother, a filter and a forward predictor of the input values. When input signal confidence is low, the mode parameter is adjusted so that the processor smoothes the input signal. When input signal confidence is high, the mode parameter is adjusted so that the processor output has a faster and more accurate response to the input signal. The variable mode averager is particular applicable to the monitoring of critical physiological parameters in patient-care settings. When applied to pulse oximeter oxygen saturation measurements, the mode parameter can be varied in real-time to achieve a tradeoff between the suppression of false alarms and signal artifacts and the immediate detection of life threatening oxygen desaturation events.
-
Citations
19 Claims
-
1. A variable mode averager comprising:
-
a buffer that stores a plurality of weighted input values;
a mode input that specifies a time value relative to said input values; and
a processor coupled to said buffer, said processor configured to provide an estimate of said input values, said estimate corresponding to said time value. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9)
-
-
10. A signal averaging method comprising the steps of:
-
identifying a plurality of signal values;
determining a plurality of weights corresponding to said signal values;
computing a trend of said signal values adjusted by said weights;
specifying a time value relative to said signal values based upon a characteristic associated with said signal values; and
estimating said signal values based upon said trend evaluated at said time value. - View Dependent Claims (11, 12, 13)
determining a confidence level associated with said signal values; and
specifying said time value based upon said confidence level.
-
-
12. The signal averaging method of claim 11 wherein said trend is a linear LMS fit to said signal values adjusted by said weights.
-
13. The signal averaging method of claim 12 wherein said time value generally corresponds to the maximum time of said signal values when said confidence level is high and said time value generally corresponds to the weighted average time of said signal values when said confidence level is low.
-
14. A signal averaging method comprising the steps of:
-
providing an input signal;
setting a mode between a first mode value and a second mode value; and
generating an output signal from an estimate of said input signal as a function of said mode, wherein said output signal generally smoothes said input signal when said mode is proximate said first mode value and said output signal generally tracks said input signal when said mode is proximate said second mode value. - View Dependent Claims (15, 16, 17, 18)
setting said mode proximate said first mode value when said confidence level is low; and
setting said mode proximate said second mode value when said confidence level is high.
-
-
18. The signal averaging method of claim 14 wherein said input signal is a physiological measurement and said setting step comprises setting said mode proximate said first mode value when said measurement is corrupted with noise or signal artifacts, otherwise setting said mode proximate said second mode value so that said output signal has a fast response to physiological events.
-
19. A signal averager comprising:
-
an input means for storing a plurality of signal values;
an adjustment means for modifying said signal values with a plurality of corresponding weights;
a curve fitting means for determining a trend of said signal values;
an estimate means for generating an output value along said trend; and
a mode means coupled to said estimate means for variably determining a time value at which to generate said output value.
-
Specification