Pulse oximeter sensor with piece-wise function

IS 2 392 818 Τ3

DESCRIPTION

PULSE OXIMETER SENSOR WITH PIECE FUNCTION

BACKGROUND OF THE INVENTION

[0001] The present invention relates to memory oximeter sensors.

[0002] Pulse oximetry is commonly used to measure various characteristics of blood flow including, but not limited to, blood oxygen saturation of hemoglobin in arterial blood, and the rhythm of blood pulsations corresponding to the heart rate of a patient. . The measurement of these characteristics has been achieved through the use of a non-invasive sensor that passes light through a part of the patient's tissue where the blood perfuses the tissue, and photoelectrically detects the absorption of light in said tissue. The amount of light absorbed is then used to calculate the amount of the blood component being measured.

[0003] The light that passes through the tissue is selected to be of one or more wavelengths that are absorbed by the blood in an amount representative of the amount of the blood component present in the blood. The amount of reflected or transmitted light passed through the tissue will vary according to the changing amount of the blood component in the tissue and the related light absorption. To measure the oxygen level in blood, said sensors have been provided with light sources and photodetectors that are adapted to operate at two different wavelengths, according to known techniques for measuring oxygen saturation in blood.

[0004] Various methods have been proposed in the past for encoding information in sensors, including pulse oximeter sensors, to transmit useful information to a monitor. For example, an encryption mechanism is shown in Nellcor's U.S. Patent No.^s 4,700,708. This mechanism refers to an optical oximetry probe that uses a pair of light-emitting diodes (LEDs) to direct light through tissue perfused by blood, with a detector that collects light that has not been absorbed by the tissue. . The operation depends on the knowledge of the wavelength of the LEDs. Since the wavelength of the LEDs can vary from one device to another, a coding resistor is placed on the sensor, the value of the resistor being the one corresponding to the actual wavelength of at least one of the LEDs. When the oximeter instrument is turned on,

ES 2 392 818 Τ3 first determines the resistance value and thus the suitable saturation calculation coefficients for the value of the wavelengths of the LEDs in the probe. [0005] Other encryption mechanisms have been proposed in US patents n^and 5.259.381, 4.942.877, 4.446.715, 3.790.910, 4.303.984,

4,621,643, 5,246,003, 3,720,177, 4,684,245, 5,645,059, 5,058,588, 4,858,615 and 4,942,877. The '877 patent in particular discloses the storage of a variety of data in a memory of a pulse oximetry sensor, including the coefficients for a saturation equation for oximetry.

[0006] Nellcor pulse oximeter sensors are encoded with a resistance value (RCAL) that corresponds to the wavelength or wavelengths of the LED / -S in the emitter, as described in patent n^and 4,700,708. Nellcor pulse oximeter instruments read this resistance coding value and use it as a pointer to a look-up table containing the appropriate set of coefficients for that sensor to calculate arterial oxygen saturation (SpC> 2). The function that converts the modulation ratio R (also known as the ratio of ratios or rat-rat) of the measured IR and red signal into a calculated saturation value is derived from the basic form of the Lambert-Beer law:

5'Α »* (Ι-ί) 'Λ where I] and I2 refer to the light signals detected at two different points in the cardiac cycle, and β₅ refers to the characteristic light absorption properties of oxygenated and deoxygenated hemoglobin. When solved for saturation (S), the result takes the following form:

-5100 = c¡ »-c_tR (c, + -c,) • 100 · (2)

[0007] Equation 2 can be further simplified to require only three constants (eg by dividing each constant by C2), but will be used as shown for the remainder of the description. Although based on theory, the four constants c- | - C4 are determined empirically. The theoretical values ​​of the constants are insufficient mainly due to the dispersion complexity of

ES 2 392 818 T3 light and sensor optics. The values ​​of the constant sets (c- | through C4) vary with each resistor coding bin (each bin corresponds to a range of different characterized LED wavelengths). Multiple sets of coefficients (bins) are provided in a look-up table on Nellcor oximeters. When SpC values> 2 calculated according to Eq. 2 are less than 70%, a revised SpC value> 2 is used using a linear function:

SpO_t = f, -c_to Λ, (3) where both C3 and οθ vary with the encoding value of the resistor. This linear function has been found to better match spÜ2 (the arterial oxygen saturation measured by the pulse oximeter) with SaO2 (the actual arterial oxygen saturation value, measured directly in a blood sample) in observation performed at low saturations.

[0008] A limitation of this method is that the correct calibration of the pulse oximeter sensor can only be achieved if the relationship between the modulation ratio of the signal (R) and the SaC> 2 of the blood is adjusted to one of the sets of precoded calibration coefficients.

[0009] Another limitation of this method is that the relationship between R and SaOy of the pulse oximeter sensor may not be linear in a region of low saturation, or that the cut-off point may not be optimally located at 70% SpO2 [ 0010] Yet another limitation of the prior art method is that the functional relationship between true arterial oxygen saturation and the measured signals may not fit a single function over the full spectrum of the measurement range.

[0011] A system according to the invention is defined in claim 1. Some preferred features are defined in the dependent claims.

SUMMARY OF THE INVENTION

[0012] The present invention takes advantage of a memory in the sensor to provide improved performance. Multiple sets of coefficients are stored. Multiple sets are applied to different ranges of saturation values ​​to provide a better fit by breaking the ratio of R to SpO₂ in different pieces,

ES 2 392 818 T3 each described by a different function. The different functions can also be respective formulas for determining oxygen saturation.

The sensor also stores a variable between the two functions used for oxygen saturation. The two functions could be either separate formulas or the same formula with different coefficients. This allows optimization to a value other than the 70% cut-off value of the prior art.

[0014] In another aspect of the present invention, the sensor can store more than one cutoff point to create more than two functions that describe the ratio of R to SpO2.

[0015] In yet another aspect of the present invention, a spline function is used, cutting the ratio of R to SpO2 in an arbitrary number of regions.

[0016] Each of the described methods improves the fit between the selected mathematical function and the arterial oxygen saturation by cutting the relationship into subsets of the entire measured range and determining the optimal coefficients for each range. Spline fitting, in this context, similarly cuts the entire measurement range into subsets to efficiently describe the numerical relationship between the underlying tissue parameter of interest and the actual signals that are used to estimate this value.

[0017] For a better understanding of the nature and advantages of the invention, reference will be made to the following description in conjunction with the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

[0018]

Fig. 1 is a block diagram of a pulse oximeter system incorporating the present invention.

Fig. 2 is a plot of R (signal modulation ratio) versus oxygen saturation (SaO2).

Fig. 3 is a diagram of the contents of a sensor memory according to the invention.

Fig. 4 is a graph of oxygen saturation versus R to illustrate the embodiment of spline or curve fitting to a predefined set of knots.

Figures 5A, 5B, 6A and 6B are graphs illustrating improved curve fit of embodiments of the invention versus prior art.

DESCRIPTION OF THE SPECIFIC IMPLEMENTATION MODES

Sensor monitor / reader

ES 2 392 818 T3

Fig. 1 is a block diagram of an embodiment of the invention. Fig. 1 shows a pulse oximeter 17 (or sensor reader) that is connected to a non-invasive sensor 15 attached to the tissue of the patient 18. The light from the LEDs 14 of the sensor passes through the tissue of the patient 18 and after being transmitted to Light is received through or reflected from tissue 18 by a photosensor 16. Two or more LEDs may be used depending on the embodiment of the present invention. The photosensor 16 converts the received energy into an electrical signal, which is then input to the input amplifier 20.

[0020] Light sources other than LEDs can be used. For example, lasers can be used, or a white light source with filters of suitable wavelength could be used either at the transmitting end or at the receiving end.

[0021] The Time Processing Unit (TPU) 48 sends control signals to the LED unit 32, to activate the LEDs, usually alternately. Again, depending on the embodiment, the unit can control two or any additional number of desired LEDs.

The signal received from the input amplifier 20 is passed through two different channels shown in the embodiment of Fig. 1 for two different wavelengths. Alternatively, three channels could be used for three different wavelengths, or N channels for N wavelengths. Each channel includes an analog switch 40, a low-pass filter 42, and an analog-to-digital (A / D) converter 38. The control lines of the TPU 48 select the appropriate channel at the time that the corresponding LED 14 is being driven, in synchronization. A queued serial module (QSM) 46 receives digital data from each of the channels via data lines 79. CPU 50 transfers the data from QSM 46 to RAM 52 as it proceeds. the QSM 46 fills up periodically. In one embodiment, the QSM 46, TPU 48, CPU 50, and RAM 52 are part of an integrated circuit, such as a microcontroller.

Sensor memory

Sensor 15, which includes photodetector 16 and LEDs 14, has a sensor memory 12 associated therewith. The memory 12 is connected to the CPU 50 in the sensor reader or monitor 17. The memory 12 could be packaged in a sensor body 15 or in an electrical outlet connected to the sensor.

[0024] Fig. 2 is an example of a graph of the ratio of ratios (R) on the X axis versus oxygen saturation (SaÜ2) on the Y axis. A cut-off point 52 is shown.

ES 2 392 818 T3 prior art, the cut-off point of 70% was predefined in the monitor software. To the right of the cut-off point (oxygen saturations between 70-100%) a formula with four coefficients was used. To the left of the cutoff point in the prior art, a linear equation with two coefficients was used. The present invention provides greater flexibility and accuracy using a non-linear formula for the part of the curve to the left of the cutoff point 52. By using a memory chip in the sensor itself, it is possible to store these coefficients on the memory chip, as well as separate coefficients for higher saturation values. [0025] In another embodiment of the invention, the cutoff point 52 can be stored in the memory chip, and can be chosen to optimize the curve fit for the two sets of coefficients. In other words, a better fit to the two curves can be obtained if the cut-off point is 68%, for example. In an alternative embodiment, multiple cut points and curves can be used. Also, instead of using the same formula, different formulas can be used for the different sections in another embodiment.

[0026] Fig. 3 illustrates the contents of the memory of sensor 12 of Fig. 1. As shown, a first set of coefficients is stored in a first section of memory 54. A second memory portion 56 stores a second set of coefficients. Finally, in a third memory section 58, the cut-off point 52 is stored. Different combinations of these elements can be stored in different memories. For example, the cut point could be excluded for some, and in others a cut point could be provided with only one set of coefficients (with the other set of coefficients on the monitor). Alternatively, the cut-off point can be determined by a sensor model number that is stored in memory, or some other identifying value.

Equation β:

[0027] In one embodiment, an improved form of curvilinear function is used. Instead of using Eq. 3 (linear) in the lower saturation region, Eq. 2 (non-linear) is used for both the upper and lower saturation regions. The cutoff point that defines when to change the coefficients from a set in the upper region to a set in the lower region is defined by another coefficient. The cut-off point can be programmed either as a value of R, or as a value of SpC> 2. With the cut-off point defined as a value of R, the algorithm becomes:

IS 2 392 818 Τ3

SpO₂ bdR (cd) -R + (ba)

<img file="ES2392818T3_D0001.tif" />

- c ^, b - c₂,DC₃, dc₄⁼ fe&gt; b - C-], cc₃, d - Cg (4)

Curve fit

[0028] Multiple region curve fitting follows the same methodology as single region fit. Simply put, the data is divided into separate regions and the coefficients for each region are determined separately. Software programs are available on the market, (eg Mathcad, (Mathsoft, Inc, Cambridge, MA). The process can also be found in, for example, Data Reduction and Error Analysis for the Physical Sciences (Philip Beviyton, McGraw-Hill, New York 1969, Ch. 11 - Least squares fit to an arbitraty function).

Spline adjustment

[0029] An alternative embodiment uses a spline (curve) fit, or higher or linear order interpolation to a predefined set of SpC&gt; 2 versus R values ​​(knots) · A knot is a term of the art in the spline fit that refers to an xy pair that corresponds to a node on a line, with a number of such nodes defining the line. Spline adjustment is an interpolation technique.

[0030] For example, R values ​​at specifically defined values ​​of SpC&gt; 2 are stored in sensor memory. This would be an example of the above:

R = abc

SpC&gt; 2 = 100 9590

Alternatively, although less preferably, the independent variable could be swapped:

R = 0.5 0.60.7

SpO₂ = x and z

a) Only bold values ​​(e.g. a, b and c) need to be stored with preselected and fixed spaced values ​​of SpC&gt; 2 (evenly spaced or not). Or, alternatively, preselected values ​​of R.

b) An alternative method stores in the sensor memory the values ​​of

ES 2 392 818 T3

SpC&gt; 2 (minimum) and SpC&gt; 2 (maximum) of the spline margin, the number of nodes that will be defined, and the sequence of defined values ​​of R for these nodes.

c) Another alternative method could store the SpO2 and associated R value for each node.

For each of these options, the instrument would use a spline fit algorithm, preferably a cubic spline, to determine the SpÜ2 at the measured R value based on the stored values ​​(an alternative could be a linear or higher order interpolation algorithm ).

[0031] Fig. 4 illustrates the cubic spline method. Fig. 4 is a graph of oxygen saturation versus R for a particular sensor emitter. Therefore, instead of storing the coefficients as in the prior art method, the actual oxygen saturation and R values ​​are calculated and stored in sensor memory for the specific sensor characteristics (e.g., emitter wavelengths). When the oximeter measures the signal level of the light detector, it determines an oxygen saturation value by determining the point on the curve associated with the R value calculated between two of the sample points shown in Fig. 4.

[0032] There is a trade-off between the number of defined nodes and the amount of memory required to store them. Very few nodes require very little storage memory, but may not adequately describe the functional relationship; too many over-define the curve and consume more memory. The inventors have found that knots spaced 5% -10% give adequate results.

Cubic spline calculation:

[0033] The cubic spline interpolation process is known to those of skill in the art. Intrinsic to using the spline method is that the value of R needs to be determined first before being translated into SpO2- The preferred spline interpolation process can be accomplished using the functions provided in Mathcad, and it treats the end points with cubic functions. Other references are available for cubic spline interpolations.

[0034] The process of finding the coordinates of the nodes in empirical data with a significant amount of noise may require an additional step. The basic curve fitting programs available on the market can be used (sigmaPlot, or

ES 2 392 818 T3

TableCurve, or Mathematical for example) to determine the best-fit functional approximation of the data. Alternatively, one can perform a least squares fit of an arbitrarily selected analytic function and choose the R values ​​at the node locations (SaO2 values). The analytic function can be an overlapping piecewise polynomial (eg, linear or parabolic), or the curvilinear equation of Eq. 1 or Eq. 4. Another approach is to carry out a least squares selection of the nodes directly.

[0035] Figure 5A shows the conventional curve fit of the prior art, in which a linear relationship is used below 70% saturation, with a curvilinear focus above 70%. The residual error due to the imperfect fit of the true R for the SaO2 response of the curvilinear approach above 70% saturation is illustrated by curve 60, while the residual error of the linear interpolation approach below 70% is shown below. Shine through dots 62. Figure 5B illustrates the use of curvilinear fits in both regions, using a different curvilinear curve 64 below 70%. In this example, a much improved fit is provided. In both figures, the smallest dotted line 66 corresponds to the use of a single curvilinear fit along both regions, which is also not so exact, having a much higher error characteristic compared to the curves of the invention, 64 and 60 of Figure 5B.

[0036] Figure 6A and 6B show a plurality of knots as circles 70 in the graphs. Dotted line 72 in Figure 6A illustrates a linear interpolation fit to these nodes, showing a result prone to residual error with multiple curves. In Figure 6B, on the other hand, the present invention using a cubic spline fitting method provides a dotted line 74 that exhibits a more exact fit at knots 70.

As those skilled in the art will understand, the present invention can be practiced in other specific embodiments without departing from the essential characteristics thereof. For example, any function can be used for the formulas to determine oxygen saturation, not just those described. For limited sensor memory, the function representation can be compressed. Any representation of a function can be used. The calibration coefficients can be based on more or different characteristics in addition to the wavelength (s) of the sensor LEDs. For example, other characteristics of the LED emitter or characteristics of the sensor design may be factors in the sensor's calibration coefficients.

Furthermore, the formula for calculating oxygen saturation can be a

ES 2 392 818 T3 function of more than the ratio of ratios; for example, other input variables such as signal strength, light levels, and signals from multiple detectors could be used. [0039] The foregoing description is intended to be illustrative, but not limiting, of the scope of the invention that is set forth in the following claims.