Spectral analysis for a more reliable determination of physiological parameters
First Claim
1. An apparatus for determining a spectral ratio between a first signal comprising a first spectrum which depends on a biological quantity, and a second signal comprising a second spectrum which depends on a biological quantity, the first signal and the second signal comprising a plurality of harmonics of a periodic signal, the apparatus comprising:
- a computer for computing a first wave ratio between a spectral value of the first spectrum at a first frequency which comprises a harmonic of the plurality of harmonics, and a spectral value of the second spectrum at the first frequency;
and for computing a second wave ratio between a spectral value of the first spectrum at a second frequency which comprises another harmonic of the plurality of harmonics, and a spectral value of the second spectrum at the second frequency;
an optimizer for determining the spectral ratio while using the first and second wave ratios, the optimizer being adapted to determine the spectral ratio such that it differs from the first wave ratio and the second wave ratio, and meets an optimization criterion,wherein the computer is adapted to pass on a plurality of the spectral values of the two spectra to the optimizer as tuples of a frequency,wherein the spectral values are complex spectral values,wherein the optimizer is adapted to determine a regression line from the plurality of tuples of the complex spectral values, the actual or perpendicular square distance from the tuples to the regression line being minimized as the optimization criterion, andwherein the optimizer is adapted to determine the regression line by applying the total least squares fit method, and the spectral ratio depends on the regression line.
1 Assignment
0 Petitions
Accused Products
Abstract
An apparatus for determining a spectral ratio between a first signal having a first spectrum which depends on a biological quantity, and a second signal having a second spectrum which depends on a biological quantity, the first signal and the second signal having a plurality of harmonics of a periodic signal, the apparatus including a computer for computing a first wave ratio between a spectral value of the first spectrum which has a frequency of a harmonic of the plurality of harmonics, and a spectral value of the second spectrum which has the frequency of the harmonic; and for computing a second wave ratio between a spectral value of the first spectrum which has a frequency of another harmonic of the plurality of harmonics, and a spectral value of the second spectrum which has the frequency of the other harmonic. In addition, the apparatus includes an optimizer.
-
Citations
25 Claims
-
1. An apparatus for determining a spectral ratio between a first signal comprising a first spectrum which depends on a biological quantity, and a second signal comprising a second spectrum which depends on a biological quantity, the first signal and the second signal comprising a plurality of harmonics of a periodic signal, the apparatus comprising:
-
a computer for computing a first wave ratio between a spectral value of the first spectrum at a first frequency which comprises a harmonic of the plurality of harmonics, and a spectral value of the second spectrum at the first frequency; and for computing a second wave ratio between a spectral value of the first spectrum at a second frequency which comprises another harmonic of the plurality of harmonics, and a spectral value of the second spectrum at the second frequency; an optimizer for determining the spectral ratio while using the first and second wave ratios, the optimizer being adapted to determine the spectral ratio such that it differs from the first wave ratio and the second wave ratio, and meets an optimization criterion, wherein the computer is adapted to pass on a plurality of the spectral values of the two spectra to the optimizer as tuples of a frequency, wherein the spectral values are complex spectral values, wherein the optimizer is adapted to determine a regression line from the plurality of tuples of the complex spectral values, the actual or perpendicular square distance from the tuples to the regression line being minimized as the optimization criterion, and wherein the optimizer is adapted to determine the regression line by applying the total least squares fit method, and the spectral ratio depends on the regression line. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24)
-
-
25. A method for determining a spectral ratio between a first signal comprising a first spectrum which depends on a biological quantity, and a second signal comprising a second spectrum which depends on a biological quantity, the first signal and the second signal comprising a plurality of harmonics of a periodic signal, the method comprising:
-
computing a first wave ratio between a spectral value of the first spectrum at a first frequency which comprises a harmonic of the plurality of harmonics, and a spectral value of the second spectrum at the first frequency; computing a second wave ratio between a spectral value of the first spectrum at a second frequency which comprises another harmonic of the plurality of harmonics, and a spectral value of the second spectrum at the second frequency; determining the spectral ratio while using the first and second wave ratios, the spectral ratio being determined such that it differs from the first and second wave ratios and meets an optimization criterion, wherein the step of determining comprises; passing on a plurality of the spectral values of the two spectra as tuples of a frequency, and wherein the spectral values are complex spectral values, computing a regression line from the plurality of tuples of the complex spectral values, the actual or perpendicular square distance from the tuples to the regression line being minimized as the optimization criterion, wherein the regression line is determined in accordance with the total least squares fit method, and the spectral ratio depends on the regression line.
-
Specification