DE3234388A1 - Method and device for quantitative determination of the oxygen saturation in the blood from photometric measured values - Google Patents

Abstract

The invention relates to a method and a device for quantitative determination of the oxygen saturation in the blood from photometric measured values, which are obtained periodically by means of a photometer with a monochromator. Hitherto, essentially only qualitative conclusions could be drawn from photometric spectra. According to the invention, a first spectrum is taken from the periodic signal sequence, a predetermined number of additional spectra are chosen by shape comparison, a Fourier transformation of the spectra is carried out in the frequency domain and, on the basis of the amplitudes in the frequency domain, a calibration curve of at least two measurement samples is determined. The related device with an evaluation circuit comprises in this case at least one evaluation unit, a shape comparison unit, a unit for Fourier transformation and a calibration device. Routine use is therefore possible. <IMAGE>

Description

Method and device for quantitative determination

the blood oxygen saturation from photometric measurement values The invention relates to a method and an apparatus for quantitative determination the blood oxygen saturation from photometric measured values, which by means of a Photometer with monochromator can be obtained periodically.

The degree of oxygenation plays a role in the supply of oxygen to tissues the oxygen carrier hemoglobin in the capillary network plays a decisive role. With help reflection spectroscopy can be due to the different absorption behavior of oxygenated and reduced hemoglobin is in principle a statement about the momentary Degree of oxygenation in the superficial capillaries are taken.

It is already known to carry out reflection spectroscopy "in vivo". The examination is particularly carried out through the use of flexible light guides possible in small areas, with artifacts caused by movements of the tissue surfaces should be prevented by short recording times.

A fiber optic photometer has already been proposed using which determines a signal spectrum of a tissue sample as a function of the wavelength can be. An expert can use the amplitude curve of the determined spectrum Derive statements about the oxygenation of the tissue sample.

However, the predictions remained largely with the previously known method limited to a qualitative interpretation of the spectra. It is a matter of invention dat £ r 'to specify a method and to create an associated device with those quantitative statements can be made.

The object is achieved according to the invention in that from the periodic Signal sequence a first spectrum is picked out, a predetermined one by shape comparison Number of further spectra is selected, a Fourier transform of the spectra is made in the frequency space and based on the amplitudes in the frequency space a Calibration curve is determined from at least two test samples.

The associated device with an evaluation circuit includes at least one selection unit, a shape comparison unit, a unit for Fourier transformation and a calibration device.

In an advantageous further development, the selection is made to be the most similar the determined spectra subjected to interference suppression; this is preferably done by adding and averaging (so-called averaging). The number of spectra needed can be limited to a reasonable level.

The invention can now be used in a surprisingly simple manner evaluate the periodically occurring spectra. After calibration has been carried out, make precise quantitative statements about the degree of oxygenation of any test samples. This now shows a way with which the use of the method described is made much easier in practice.

Further advantages and details of the invention emerge from the following description of the figures based on the drawing in connection with the other subclaims.

They show: FIG. 1 the schematic structure of the photometer with spectrometer, FIG. 2 shows an original curve that was measured with the spectrometer used, FIG. 3 the result of a selective averaging, FIG. 4 the associated curve in the frequency domain, 5 shows such curves of two calibration samples, one deoxygenated as calibration samples and an oxygenated sample can be used, and FIG. 6 shows a block diagram Computing circuit for carrying out the method steps illustrated with reference to FIGS. 2 to 5 In FIG. 1, a micro-fiber optic photometer consists essentially of one Light source 1, which is realized by a xenon high pressure lamp. From the light source 1, the light is passed through a filter system 2 onto an optical fiber 3 with a diameter of 70 μm focused as a transmitting light guide and passed over it to the measuring point of a tissue sample P. The reflected light is transmitted via six fibers that are ring-shaped around the transmitting light guide are arranged and generally denoted by 4, led to a wavelength selector. The latter consists of a progressive interference filter disc rotating by means of a Moter @@ 5 as a monochromator, where the light varies depending on the angle of rotation Wavelength between 495 nm and 615 nm is selected. The interference filter disc 6 leaves the light different at the measuring point M depending on the angle of rotation Wavelengths passier with a rotation angle of G to 1800 of the wavelength range between 495 and 615 nm in ascending order and in the range from 180 to 3600 in is traversed in the opposite direction. The intensity of the transmitted light is measured by a photomultiplier 7 and in the form of an analog voltage signal issued for the purpose of further settlement. The voltage signal is sent to a Evaluation device 10. With each disk revolution of the monochromator 5, a Trigger pulse for the evaluation can be derived.

It is also possible to use a spectrometer in which the irradiated Light is monochromatized and then the spectrum is scanned periodically.

The evaluation device is described in detail with reference to FIG. The procedural treatment essentially emerges from FIGS. 2 to 5 of the measurement signals.

FIG. 2 shows, on the one hand, that the measurement signal is periodic with the rotation the interference filter disc 5 runs and on the other hand that it is of a strong Noise is superimposed. For further signal processing, the signals in separate periods. The difficulty here is that on the one hand the period length is unknown and on the other hand the lengths of different periods are not identical.

The latter can, for example, be caused by fluctuations in the speed of rotation the progressive interference filter disk 5 are effected.

In the proposed method, a reference period " determined and found the most similar sections by means of correlation. Here can be used that the interference filter disc 5 is already independent of the measurement signal generates a periodic basic function, whose maxima or minima for period determination to serve.

By correlating individual periods with the "reference period" found the most similar sections. The correlation length is in fact Ii'ciien determined by the length of the reference spectrum; but it can be different Compensate for period lengths by overlapping the correlation fields.

With the described process sub-steps you are off the trigger pulse the filter disc independently; the possibility arises from the large number from existing spectra a certain number of individual spectra based on quality criteria to select. Ice such a quality criterion can, for example, be given by 'Form comparison determined correlation coefficient.

A flow suppression can in general by a weighted Averaging can be achieved. One can show that an averaging over 30 to 60 periods already leads to a good result.

For example, a frequency spectrum is shown in Figure 3, evaluated according to the method described above and averaged over Has undergone 50 periods.

Since the input signals are periodic and band-limited, they can can be transformed in the frequency domain with the help of the Fourier transformation. The transformation leads to a representation of the function as an amplitude frequency diagram and is shown in FIG. 4.

Changes in hemoglobin oxygenation cause a significant change of the corresponding amplitude-frequency diagrams. It has been shown that small Changes in the oxygenation admittedly only comparatively small changes in the signal curve cause which differences are hardly or not noticeable visually; in the frequency domain However, there is a clear difference.

FIG. 5 now shows such spectra, on the one hand with a deoxygenated calibration sample and, on the other hand, obtained with an oxygenated calibration sample became.

It is useful for spectra that are under the same conditions without Changes in concentration were recorded to determine the following parameters: a) m (t): corresponds to a noise reduction through averaging b) h (t): corresponds to Calibration sample with the most strongly oxygenated spectrum c) 1 (t): corresponds to a Calibration sample with the most strongly deoxygenated spectrum. The proportion from h (t) to m (t) based on the following relationship: m (t) = α h (t) + #. 1 (t) (1) The factor <represents a measure for the oxygenated and a measure of the deoxygenated part of the spectrum.

In particular, the factor α of interest can now be determined in this way that equation (1) both with the spectrum of a fully oxygenated calibration sample and the spectrum of a fully deoxygenated calibration sample is correlated. It leaves a linear calibration curve can thus be represented, from which the degree of oxygenation is arbitrary Samples can be determined. This is a quantitative statement about the degree of oxygenation given.

In FIG. 6, A denotes the analog signal of the photomultiplier 7. This is via a filter 11 for band limitation and an A / D converter 12 with high sampling frequency converted digitally and then a memory 15 @ for the purpose Intermediate storage carried out. Via a trigger signal generator 15, a pulse can are guided from the rotation of the monochromator disk 5. In the present case a trigger pulse is used in particular to select the first reference spectrum. In a comparison unit 16 t this "reference period" with the in memory 15 cached spectra compared. From the large number of stored A predetermined number of further spectra is thus obtained by comparing the shape of the measured values determined, which together the averaging unit 20 for the purpose of interference suppression ' are fed. After addition and averaging, the signal is largely Noise-free, so that it was fed to a unit 25 for Fourier transformation can. A signal selector 26 selects from the Amp, itu1en frequency diagram for the Oxygenation shows significant frequencies. The signal selector 26 is one Calibration facility 27 assigned. It can be done by entering at least two calibration samples with known Determine a characteristic curve for oxygenation. By comparison with the characteristic in Per Unit 28 then determines the oxygenation of unknown samples and on a display unit 29 digital or analog; e displays.

The described method and the described device sml for quantitative determination of the blood oxygen saturation of tissues with constant Hemoglobin content suitable, In addition, the proposed measurement method can also be used at different hemoglobin concentrations.

13 claims 6 figures

Claims (13)

Claims 9 method for the quantitative determination of the ¢ ut oxygen saturation from photometric measured values, which are periodically measured by means of a photometer with a monochmator be obtained, d a d u r c h g e -k e n n n z e i c h n e t that a) from the periodic Signal sequence a first spectrum is picked out, b) a shape comparison predetermined number of further spectra is selected, c) a Fourier transformation the spectra is made in the frequency domain and d) based on the amplitudes im Frequency domain a calibration curve is determined from at least two test samples. 2. The method of claim 1, d a d u r c h g e -k e n n z e i c h n e t that for method step a) the first spectrum by means of a trigger pulse is derived. 3. The method of claim 1, d a d u r c h g e -k e n n z e i c h n e t that the shape comparison is carried out after digitizing the spectrum data will. 4. The method of claim 1, d a d u r c h g e -k e n n z e i G h n e t that the measurement signal is additionally subjected to an interference reduction. 5. The method according to claim 4, d a d u r c h g e -k e n n z e i c h n e t that the interference reduction by means of selective addition and averaging takes place. 6. The method according to claim 5, d a d u r c h ge e n n z e i c h n e t that for the selective addition between 20 and 60 spectra, preferably; 50 spectra can be selected. @@@ 7. The method according to claim 1, d a d u r c h G e -k e n n n z e i c h n e t that for process step @ d) a first deoxygenated Measurement sample and a second oxygenated measurement sample can be used. 8. Apparatus for performing the method according to claim 1 or one of claims 2 to 5, characterized in that the evaluation circuit at least a selection unit (11-15), a shape comparison unit (16), a * for Fourier transformation (25) and a device (26-28). 9. Apparatus according to claim 8, d a d u r c h & g ge -k e n n notices that an additional unit for interference reduction (20) is available is.- 10. The device according to claim 9, d a by k e n n z e i c h n e t that the The unit for reducing interference is an averaging unit (20). 11. The device according to claim 8, characterized in that it is e t that a memory (15) is present in which the recorded spectra for be cached for further processing. 12. The device according to claim 8, characterized in that it is not possible to that a digital computer is available for evaluation. 13. The apparatus of claim 12, d a d u r c h g e k e n n z e i c h n e t that a Nikroprocessor with associated memory is used as a digital computer will.