JP3124073B2 - Blood oxygen saturation monitor - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to an monitor equipment that may be used to measure oxygen saturation of the living arterial, especially, to measure the blood oxygen saturation that is not affected by signal artifacts
That on the equipment.

[0002]

2. Description of the Related Art Hemoglobin and oxyhemoglobin exhibit different light absorption spectra, and an optical oximeter is constructed by utilizing this difference in the absorption spectrum as a basic principle. Most currently available oximeters that measure blood oxygen saturation by optical techniques are of the transmission type. These devices transmit light through a protruding part of a living body such as a finger or an earlobe, and compare the light transmitted on one side of the protruding part with the light detected on the other side, thereby obtaining an oxygen saturation level. To determine. However, a major problem with transmission-type measurement methods is that they can only be used on living parts that are thin enough to allow light transmission. In recent years, attention has been focused on the development of a reflective oximeter that measures blood oxygen saturation using reflected light. The reflection oximeter is advantageous in that it can measure blood oxygen saturation even in a living body part that is not suitable for transmission-type measurement.

Various methods and devices that utilize the optical properties of blood to measure blood oxygen saturation have been disclosed in prior patent literature. Representative devices employing transmission-type measurement methods are described, for example, in US Pat.
No. 446,871, No. 4,407,290, No. 4,2
No. 26,554, No. 4,167,331, No. 3,99
It looks like 8,550. Also, reflective measuring devices and techniques are described, for example, in US Pat. No. 4,447,150,
Nos. 4,086,915 and 3,825,342. Theoretical considerations on the design principle of a reflective oximeter are given by Yitzhak
Mendelson), "Theory and Development of a Percutaneous Reflex Oximeter Device for Non-Invasive Measurement of Arterial Oxygen Saturation" (Published Doctoral Dissertation, No. 8329355, University Microfilms, Ann Arbor, Mich., 19
1983). Also, theoretical considerations on the optical properties of blood are given by Narayanan R. Narayanan R. Pisharoty, “optical scattering by blood” (published doctoral dissertation, No. 7
No. 124816, University Microfilms, Ann Arbor, Michigan, 1971).
In addition, many theoretical studies have been disclosed which analyze the behavior of light in blood and other samples. A brief introduction to some of them is Paul Kubelka.
In "Studies on Optical Properties of Samples that Strongly Scatter Light-Part 1" (Journal of the Optical Society of America, Vol. 38, No. 5)
No., May, 1948); J. RJZdrojkowski and N.J. R. Pisharoti's "Optical Transmission and Reflection by Blood" (IEEE Transaction on Biomedical Engineering, B
ME-17, No. 2, April, 1970).

[0004]

Generally, the amplitude of a pulse measured by a reflection oximeter is less than 1% of the total reflected signal intensity. Therefore, poor measurement accuracy of the pulse amplitude results in inaccurate measurement of blood oxygen saturation. A pulse amplitude measurement error is likely to occur when the measured signal waveform includes an upward or downward trend. If the detected signal waveform includes an upward trend on average, the pulse amplitude of that waveform will be smaller than it actually is. Conversely, if the detected signal waveform has an average downward trend, a downward trend will appear. , The pulse amplitude tends to be larger than it actually is. In the conventional apparatus, even when an upward or downward trend is included in the signal waveform, this is used as it is, so that the measurement accuracy of the pulse amplitude tends to decrease, and as a result, the blood oxygen saturation from those pulses is reduced. Was inaccurate. As described in detail below, equipment of the present invention solves these problems.

An object of the present invention is to provide a non-invasive oximeter that can accurately measure the blood oxygen saturation of a patient. In the present invention, the oxygen saturation of the arterial blood of the patient is determined by a non-invasive optical technique that utilizes the difference between the optical absorption spectra of hemoglobin and oxyhemoglobin.

[0006]

The most basic configuration of the present invention comprises (a) a light source for irradiating a patient's arterial blood with two kinds of light having different wavelengths, and (b) light after reaching the blood. (C) means for determining the blood oxygen saturation of the patient by correlating the light intensity with a predetermined oxygen saturation reference curve. As the wavelength of light emitted from one of the light sources, one having a substantially different absorption coefficient between hemoglobin and oxyhemoglobin is used. The optical signal detected by the present device includes an alternating current (AC) component and a direct current (DC) component of light of each wavelength. The amplitude ratio is determined based on both AC and DC components at each wavelength.
This amplitude ratio is then correlated with the oxygen saturation reference curve to determine the arterial blood oxygen saturation of the patient. Before the AD component of the optical signal is used for calculating the amplitude ratio, the effect of the upward or downward inclination is removed by the signal correction unit.

[0007]

Equipment of the operation and effect of the present invention, by excluding the influence of trends included in the signal waveform by processing the pulse signal, to solve the conventional problems related to the trend. The influence of the upward or downward trend is removed from the pulse amplitude signal used for calculating the blood oxygen saturation, thereby greatly improving the accuracy of determining the blood oxygen saturation. The trend correction method according to the present invention includes (a)
For each of the current pulse signal and the previous pulse signal,
Calculating the interval time between the diastolic blood pressure and the systolic blood pressure signal and the median pulse amplitude;
Calculating the ratio of the second amount to the second amount, and the first amount is calculated as the difference between the current median signal and the previous median signal, and the second amount is calculated as the current diastolic blood pressure-systolic blood pressure. Calculated as the difference between the inter-signal interval time and the previous interval between diastolic and systolic signals. (C) The diastolic and diastolic signals by subtracting the trend amount from the currently measured diastolic and diastolic signal values. Correcting the value, the trend amount is the product of the trend slope rate and the interval time between the diastolic blood pressure and the systolic blood pressure signal of the current pulse,
Each step is included.

[0008]

FIG. 1 shows a non-invasive blood oxygen saturation monitor 10 as an embodiment of the present invention. When the detection probe 12 is placed on the patient's tissue 14, 2
Light emitted alternately from the light emitting diodes (LEDs) 16 and 18 is reflected by the arterial blood of the tissue 14, and the reflected light is detected by the photodetector 20. In the present embodiment, the first LED 16 emits light having a wavelength of 660 nm (red light), and the second LED 18 emits light having a wavelength of 900 nm (infrared light). However, the present invention is not limited by the wavelength of light emitted from the light source. However, in order for the present invention to function effectively, the wavelength of light emitted from one of the two light sources 16 and 18 needs to have a substantially different absorption coefficient between hemoglobin and oxyhemoglobin. . The photodetector 20 outputs an electrical signal corresponding to light reflected by the arterial blood at the tissue 14 and simultaneously containing both direct current (DC) and alternating current (AC) components.

FIG. 2 is a graph showing a typical example of the waveform of a pulse signal output from the photodetector 20 for reflected light of each wavelength. FIGS. 3 and 4 show pulse amplitudes having upward and downward slopes, respectively. It is a graph which shows a signal. As noted above, these slopes impair the accuracy of measuring the pulse amplitude and thereby the accuracy of measuring blood oxygen saturation of the tissue 14. In the present invention, the influence of such a trend included in the detected signal is removed by the signal trend correcting device 22. The signal trend correction device 22 executes a trend correction method described later.

The output signal from the photodetector 20 is corrected by a trend correction device 22 and then processed by an appropriate filter 24 for light of each wavelength, thereby obtaining an A signal.
It is separated into both C and DC voltage components. The AC and DC voltage signals for each wavelength are then processed by the voltage amplitude ratio determination circuit 26 to become signals corresponding to the AC / DC ratio of each reflected light signal.
The ratio is used to determine the final composite ratio. A signal representing the composite voltage amplitude ratio is supplied to the microprocessor 28. The microprocessor 28 calculates the blood oxygen saturation according to the algorithm stored in the program memory 30 and a predetermined data reference curve. The blood oxygen saturation thus determined is displayed on an appropriate display 32.

Hereinafter, a trend correction method for determining the pulse amplitude will be described in detail. FIG. 2 is a graph showing a typical example of a pulse amplitude signal detected by the photodetector 20. This signal represents the light reflected from the tissue 14,
The time corresponding to the maximum value is the diastole of the heart (diastolic blood pressure)
On the other hand, the time corresponding to the minimum value corresponds to the systole (systolic blood pressure) of the heart.

A pulse detection window (a test window used to specify the minimum and maximum values of each pulse)
Is the section time between the diastolic blood pressure and the systolic blood pressure signal expected for the next pulse, and for example, corresponds to the section time from t3 to t4 in FIG. As described above, in the present monitor device, the maximum value and the minimum value of the previous pulse of the signal waveform transmitted through the channel from the photodetector 20 are recognized, and the maximum value (the minimum blood pressure time) of the current pulse is recognized. ), The signal waveform is not recognized as one pulse until the data sampling time reaches at least the same time as the pulse detection window. As a result, the minimum value of the signal determined in the device corresponds to the actual systolic blood pressure,
It is assured that it is not caused by artifacts such as body motion and electric noise. If the pulse detection window is too small (too short), artifacts may be erroneously recognized as pulses, while if too large (too long), two or more pulses within one window may be detected. Occurs, and eventually no pulse can be recognized.

Many conventional devices detect the interval between the diastolic and systolic signals (the amount of time between relaxation and contraction of the heart) for the pulse immediately preceding each pulse for the next occurring pulse. Used to calculate the window. For example, in FIG. 2, using the first section time t1-t2, t3-t4 is determined as the section between the diastolic and systolic blood pressure signals expected for the next pulse. Therefore, if the recognition of the minimum or systolic blood pressure time is erroneously performed in the determination of the first section time t1-t2, the detection window for the next pulse is erroneously predicted, and the pulse must be missed. There is a possibility that it will be mistakenly recognized.

Also, conventional pulse detection techniques that utilize the interval between the diastolic and systolic signals of the last recognized pulse to determine the detection window for the next pulse are susceptible to signal artifacts. . If an artifact is mistakenly identified as a pulse, the wrong diastolic-systolic interval time is used for a new pulse detection window, so that only artifacts are detected and the true pulse is missed. Easy to be. In the present embodiment, the average value of the interval time between the diastolic blood pressure and the systolic blood pressure signal for the immediately preceding four pulses is used to determine the detection window for the next pulse.

The pulse amplitude refers to an AC component of a signal waveform detected by the photodetector 20. For example, the pulse amplitude in the interval t1-t2 in FIG. 2 is a ₁ -a _2. In the present apparatus, the upper limit value and the lower limit value of the pulse amplitude are calculated using pulse amplitude information from at least one already detected pulse. The present apparatus does not recognize the pulse waveform as a pulse if the amplitude of the pulse waveform is not within the range of these upper and lower limit values. As a result, it is possible to prevent an artifact that has entered at the correct timing or a pulse that has been crushed by the artifact from being adopted as a regular pulse. If the upper and lower limit values of the pulse amplitude are very close to each other, even if they are true pulses, their amplitudes do not fall within the range, so that many correct pulses are overlooked, and the pulse amplitude If the upper and lower thresholds are far apart, false or broken pulses to be rejected will be employed.

In many conventional devices, valid pulses may be rejected because the interval between the upper and lower limits of the pulse amplitude is too narrow. In addition, incorrect pulses may be mistakenly identified as true pulses and used to determine upper and lower limits for the next pulse amplitude. On the other hand, in this embodiment, in order to calculate the upper and lower limit values of the pulse amplitude for the next pulse, the immediately preceding 4
These problems were solved by using the average value of the pulse amplitudes and the following relational expression. Lower limit value of pulse amplitude = Average value of pulse amplitude / 3.0 Upper limit value of pulse amplitude = Average value of pulse amplitude.times.3.0 3.0 of conversion factor (sale factor) used in the above equation. The numbers were selected by comparing data obtained using multiple conversion factors in the range of 1.5-5.0. The smaller the value of the conversion factor, the more rejected pulses are eliminated, and the larger the value, the more inappropriate pulses are adopted. In general, it was confirmed that when the conversion rate was about 3.0, the adoption rate of true pulses was the highest and the mixing rate of inappropriate pulses was the lowest.

In this embodiment, the program memory 3
The blood oxygen saturation (SaO ₂ ) is calculated using the algorithm stored in “0”. This algorithm determines the blood oxygen saturation based on the ratio obtained from the respective reflected light intensities of the two wavelengths of light. The following is a relational expression derived empirically. SaO ₂ = A + B × [(pulse amplitude / median signal) _red /
(Pulse amplitude / median signal) _infrared ] where A and B are constants obtained empirically, and the subscripts _red and _infrared indicate red light and infrared light, respectively.

The empirical constant A is derived mathematically, an example of which is Edgar Jr. U.S. Pat.
No. 80 and U.S. Pat. No. 4,796,636 to Branstetter et al. These contents constitute a part of the present specification. For example, the constant A can be calculated from the following equation. A = _{[(K B W Dr) / (} 35H B K Dr (W Or -
W _Rr))] + _{[W Rr / (W Rr -W Or} ) ] However, the blood volume K _B, including a scattering H _B hemoglobin by tissue and blood scattering K _Dr by blood during the diastolic blood pressure at a wavelength of 660nm change (fraction) W _Dr absorption by tissue and blood during the diastolic blood pressure at a wavelength of 660nm W _Rr is absorption light rate above formula oxyhemoglobin in absorption light ratio W _Or wavelength 660nm of the reduced hemoglobin at the wavelength 660nm Of the first term is very small, so that ignoring this does not significantly affect the accuracy of the constant A.
The second term consists solely of the absorbance of oxidized and reduced hemoglobin at a known wavelength, such as 660 nm (red). These values are known constants related to the wavelength of the light illuminating the blood. Therefore, the constant A is calculated by assigning specific numbers to the absorbances W _Rr and W _Or in the second term. For example, in the case of light of 660 nm, W
_{Since Rr} is 1.732 and W _Or is 0.211,
Substituting these values into the second term of the above equation yields a constant A of 112%.

The empirical constant B is determined empirically by calibrating the monitoring device on at least one patient having a known oxygen saturation, such as 98%.

In another embodiment, the empirical formula for calculating the blood oxygen saturation may be based on an average pulse signal instead of using a median pulse signal as described above.
Accurate results can also be obtained by calculations based on the average pulse signal, but it is not desirable to execute an averaging algorithm in microprocessor 28.

Generally, the pulse amplitude detected by photodetector 20 is less than 1% of the total reflected signal intensity. Thus, the measurement error of the pulse amplitude, even if small, makes the calculation of blood oxygen saturation inaccurate. An error in measuring the pulse amplitude can occur when the magnitude of the signal waveform includes an upward or downward slope. When the median (or average) signal of the signal waveform includes an upward slope, the pulse amplitude of the waveform becomes smaller than it actually is. The graph in FIG. 3 illustrates this,
It can be seen that the actual pulse amplitude is larger than the detected pulse amplitude. Similarly, when the median signal of the signal waveform includes a downward slope, the pulse amplitude of the waveform becomes larger than it actually is. This phenomenon is shown in the graph of FIG.

Many conventional devices use the pulse amplitude of the signal waveform as it is, even when the signal waveform includes an upward or downward slope, resulting in inaccurate results. The calculation of the degree of saturation was also inaccurate. The present monitoring device solves the above-described problem by removing the influence of an upward or downward inclination included in a signal waveform. In this embodiment,
The signal values corresponding to the lowest blood pressure and the highest blood pressure in the signal waveform are corrected as follows. First, the average time of each pulse section is determined. Section average time = (time of diastolic blood pressure + time of systolic blood pressure) / 2 Next, a median signal in each section is calculated. Median signal = (minimum blood pressure signal value + systolic blood pressure signal value) /
2. From these, the trend slope rate is calculated as follows. Trend slope rate = (current median signal−previous median signal) / (current section average signal−previous section average signal) The systolic blood pressure signal value and the diastolic blood pressure signal value are corrected as follows (however, subscripts are added). _correct and _measured indicate the correction value and the measured value, respectively. _Systolic blood pressure signal value _correct = _Systolic blood pressure signal value _measured-
(Trend slope rate * diastolic blood pressure-interval time between systolic blood pressure signals) diastolic blood pressure signal value _correct = diastolic blood pressure signal value _measured +
(Trend slope rate * distance between diastolic blood pressure and systolic blood pressure signal) Finally, the pulse amplitude is corrected as follows. Pulse amplitude _correct = diastolic blood pressure signal value _correct- systolic blood pressure signal value _correct

By removing the influence of the above-mentioned trend, the measurement accuracy of each pulse amplitude of the signal waveform is improved, and as a result, the measurement accuracy of the blood oxygen saturation by the present monitor device is significantly improved. However, when the trend change of the signal waveform is extremely sharp (for example, when only a few pulses continue), the raw data may be distorted as a result of the above trend removing operation.

One method to prevent the above problem is that each pulse should be removed because it has undergone the detrending operation, even though it would have been employed if it had not. For the pulse amplitude determined to be, the original diastolic blood pressure signal value and systolic blood pressure signal value in the "non-operated state" are adopted.
For example, despite the fact that the amplitude of the pulse subjected to the detrending operation has exceeded the upper or lower limit of the pulse amplitude,
If the pulse amplitude before the detrending operation does not exceed any of the limit values, the calculation of the blood oxygen saturation is performed based on the latter pulse amplitude not subjected to the detrending operation.

In another embodiment, the detrending method can also be performed based on information obtained from previous and / or subsequent pulses in addition to current and previous pulse information. For example, first, for the current pulse (n), the trend slope ratio between the current pulse (n) and the previous pulse (n-1) and the trend slope ratio between the current pulse (n) and the next pulse (n + 1) are separately obtained. By taking the average of the plurality of slopes or comparing them, the slope value of the overall trend is obtained, and the pulse amplitude is corrected based on that value. Further, it is possible to determine the inclination of the trend based on information from a plurality of pulses by using various curve fitting methods.

The above has been described in detail a specific example of equipment of the present invention, the present invention is not limited to any specific embodiment of the above, within the spirit and scope of the invention as set forth in the appended claims It should be understood that it includes changes, modifications and improvements.

[Brief description of the drawings]

FIG. 1 is a block diagram schematically showing a basic embodiment of a non-invasive blood oxygen saturation monitor of the present invention.

FIG. 2 is a graph showing a waveform of a pulse detected by an optical sensor of the reflection type oximeter device of the present invention.

FIG. 3 is a graph showing a waveform of a pulse including an upward trend.

FIG. 4 is a graph showing a waveform of a pulse including a downward trend.

[Explanation of symbols]

 Reference Signs List 12 detection probe 16 first light emitting diode (first light source) 18 second light emitting diode (second light source) 20 photodetector 22 signal trend correction device 24 filter 26 voltage amplitude ratio determination circuit 28 microprocessor 30 program memory

Claims (6)

(57) [Claims] A first irradiation light source for irradiating with a light of a first wavelength, a second irradiation light source for irradiating with light of a second wavelength, and a blood sample for the first and second irradiation light sources. An arrangement means for arranging for irradiation, detecting each light passing through the blood sample, converting the detected light into a signal including an AC pulsation component and a DC component, and outputting the signal; Signal correction means for correcting the AC pulsation component of each of the signals to remove the influence of an upward or downward slope from the components; and amplitude ratio calculation means for respectively obtaining the amplitude ratio of the AC component and the DC component of each of the output signals. A blood oxygen saturation monitoring apparatus comprising: means for calculating blood oxygen saturation by correlating a quotient of the two amplitude ratios with a predetermined oxygen saturation characteristic reference curve. 2. The pulse amplitude of each AC pulsation component is:
Means for rejecting the detected pulse amplitude if it falls outside a pulse amplitude range defined by a minimum and maximum pulse amplitude based on pulse amplitude information from at least one pulse amplitude detected prior to the pulse amplitude. 2. The apparatus of claim 1 , wherein the correction means includes: 3. The minimum and maximum amplitudes of the pulse amplitude range are as follows: minimum pulse amplitude = Avg _pulse ÷ conversion rate maximum pulse amplitude = Avg _pulse × conversion rate, where Avg _pulse is detected before each of the pulse amplitudes. 3. The apparatus of claim 2 , wherein the pulse amplitude is one of a pulse amplitude and an average of pulse amplitudes of at least two pulses detected before each pulse amplitude. 4. The apparatus of claim 3 wherein the conversion factor is about 3.0. 5. A diastolic blood pressure signal value and a systolic blood pressure signal value of at least one pulse detected before the pulse amplitude of each of the AC pulsation components, based on an interval time between the diastolic blood pressure and the systolic blood pressure signal. Means for setting a pulse detection window for determining a value, and only those pulses having a diastolic blood pressure signal value and a systolic blood pressure signal value separated from each other by at least the same time as the pulse detection window are suitable for calculating the blood oxygen saturation. further comprising apparatus according to claim 1 and means to certify a true pulses. 6. A section time of the pulse detection window, diastolic blood pressure for a plurality of pulses - device of claim 5 is based on the average value of the systolic blood pressure signal between the interval time.