Method and detector device for determining concentration of substance

Abstract

The invention relates to a method for determining the concentration of a substance in a sample comprising a liquid comprising particles, in particular glucose in blood, in which the refractive index of the liquid depends on the concentration of the substance dissolved therein and the particle density in the liquid is substantially constant. The method comprises obtaining a PPG first signal during a first time period, and estimating at least a portion of a noise spectrum from the first obtained signal, particularly in the frequency domain. At least a portion of the noise spectrum is used to derive a substance concentration or a change in the substance concentration.

Description

Technical Field

[0001] This application claims priority to German application DE 10 2023 130 812.2, dated November 7, 2023, the disclosure of which is incorporated herein by reference in its entirety. The present invention relates to a method for determining the concentration of a substance in a sample containing particles, particularly glucose in blood, in a liquid, wherein the refractive index of the liquid depends on the concentration of the substance dissolved therein. The invention also relates to a detector device. Background Technology

[0002] Current standard methods for measuring blood glucose typically employ an invasive technique in which a small amount of blood is drawn and then electrochemical analysis is performed using a handheld device. This method is not suitable for continuous monitoring because a fresh blood sample must be obtained by pricking the finger for each measurement.

[0003] A more recently developed technique uses a button located on the skin and employs a small needle-like sensor to miss interstitial fluid in a portion of the subcutaneous adipose tissue. However, the needle is permanently inserted into the skin. This method carries a risk of infection. Furthermore, it may be necessary to remove it in certain situations (such as during sports activities, swimming, etc.).

[0004] In addition to these invasive methods, non-invasive methods based on optical IR measurements or Raman spectroscopy also exist. In the first case, selecting a suitable emitter and detector can be difficult, while Raman spectroscopy-based methods are challenging due to their very poor signal-to-noise ratio. Recently, optical measurements using portions of the visible spectrum have been proposed, in which backscattered light is analyzed. However, the signal-to-noise ratio can vary depending on skin type, environment, and other parameters, significantly degrading the overall signal quality.

[0005] In this regard, there is a need for a method that can detect substances in liquids in a simpler way and allows for continuous measurements. Summary of the Invention

[0006] This and other objectives are addressed by the subject matter of the independent claims. Features and other aspects of the proposed principles are outlined in the dependent claims.

[0007] In addition to some other ideas based on various parameters dependent on light scattering within the tissue, this idea relies on the assessment of the noise level of the modulation depth, also known as the perfusion index. The inventors unexpectedly discovered that not only does the perfusion index itself depend on glucose concentration, but the noise itself also depends on glucose concentration. This appears to be due to the scattering introducing a glucose concentration-dependent systematic noise into the overall noise.

[0008] The perfusion index, or modulation depth, is a result of so-called photoplethysmography (PPG), which is an optical measurement of changes in blood volume (e.g., due to heartbeats) at a specific point in the body (e.g., a fingertip or other suitable location). Volume changes can be observed through signal variations during the heartbeat caused by fewer or more scatterings of different blood volumes. PPG measurements are performed over short time intervals, such as 10 seconds, or at intervals of 2 or 5 minutes.

[0009] Due to scattering and absorption of light that initially strikes the skin and subsequently may propagate through superficial and / or deep tissues (which contain vascular networks), the amount of light re-emerging at the skin-environment interface at a certain distance from the point of entry will vary depending on the optical path length (distance) traveled in the blood. This optical path length is called the "blood optical path length (BOPL)," which is directly affected by the heartbeat. Therefore, the BOPL varies over time, while the properties of the surrounding tissue portion can be considered constant, at least within the heart rate and short measurement periods.

[0010] When BOPL is at its maximum or minimum, the resulting output signal is conversely at its minimum or maximum. The difference between the minimum and maximum signals is called modulation, and the ratio of modulation to the average signal is defined as modulation depth, AC-DC ratio, or injection index.

[0011] According to Beer-Lambert's law, light absorption in blood is primarily determined by the hemoglobin present in red blood cells and the accumulated BOPL. Assuming that the density of red blood cells remains constant during measurement, it can be assumed that light nominally propagating through more or less transparent plasma will undergo scattering at the red blood cells. This means that the accumulated BOPL from the entry point to the detection point (both the entry and detection points are fixed by the device configuration) may depend on the scattering characteristics in the sense of a random walk; that is, it will become a random process.

[0012] Therefore, the perfusion index, or modulation depth, is primarily a result of direct interaction and, in particular, the scattering of light between red blood cells. This principle is currently used to measure blood oxygen saturation, also known as SpO2 measurement. SpO2 measurement is based on the ratio of PPG signals recorded at two different wavelengths (i.e., red light and IR spectra). More specifically, SpO2 measurement is performed by generating the ratio of the perfusion index at two different wavelengths. This is achieved by setting up a system in which tissue is irradiated, for example, using red light and IR emitters, for example, by alternately transmitting short light pulses at a sufficiently high repetition rate (e.g., 20 Hz, 50 Hz, 200 Hz, etc.), and sampling the corresponding photocurrent generated in a photodiode. The original photodiode current consists of an average signal level (called DC) and a signal modulation due to the heartbeat (called AC). The AC signal can be estimated, for example, by the difference between the peak and trough values ​​of the original signal.

[0013] The ratio of AC to DC is called the perfusion index (PI), or relative modulation or modulation percentage. (See reference "In Vivo Reflectance of Blood and Tissue as a Function of Light Wavelength," Weijia Cui et al., IEEE Transactions on Biomedical Engineering, Vol. 17, No. 6, June 1990). The ratio of relative modulation is calculated to estimate oxygen saturation. This process is also called the ratio of ratios. By appropriately calibrating these values ​​using a reference device, the early 100% oxygen saturation and lower levels of oxygen saturation can be identified.

[0014] However, the ratio for a given oxygen saturation and therefore for SpO2 measurement depends on the choice of optical design, particularly the relative separation between the light source and detector at each wavelength, the emission characteristics of each light source, the detection characteristics of the photodiode, and the performance of the optical barrier between the light source and detector, as well as other system aspects that affect how light can propagate from the light source to the detector. Nevertheless, the aforementioned methods for optical SpO2 measurement are widely discussed in the literature and are scientifically accepted.

[0015] It has been observed that for human subjects under steady-state conditions and with constant blood glucose levels, the perfusion index remains substantially constant. When blood glucose levels rise, the properties of plasma, and more specifically, its refractive index, change. Therefore, the difference in refractive index between red blood cells and plasma decreases, leading to reduced scattered light. Similarly, a decrease in glucose increases the difference in refractive index between red blood cells and plasma, resulting in increased scattering. This is explained in a model where scattering is primarily caused by diffraction differences; therefore, in a transparent medium where the refractive index of the liquid and the scattering particles is the same, scattering will not be observed because the light will simply propagate through the medium without interaction. However, once a refractive index mismatch exists, some light is scattered by the particles (e.g., backwards) and can therefore be distinguished from the surrounding environment (i.e., the plasma).

[0016] Therefore, the perfusion index can be used for what is known as the oral glucose tolerance test (OGTT), in which the subject starts with a baseline blood glucose level established through several hours of fasting, and then ingests a liquid or substance with a caloric value. Typically, changes in the perfusion index become visible approximately 5 to 15 minutes after ingestion.

[0017] It has been observed that the overall noise level also changes with blood glucose levels. More specifically, x-sigma, or standard deviation (a measure of noise in the remaining portion of the spectrum, i.e., the portion not modulated by the heartbeat), correlates with blood glucose concentration. Therefore, it can be inferred that noise in the PPG signal may be a carrier of additional information about the substance and its concentration in the liquid.

[0018] In particular, the proposed principle utilizes an algorithm to decompose the signal into the following components: (1) a component attributed to heartbeat modulation, and (2) a residual component, which is commonly referred to as noise and is not considered a carrier of relevant information. These components are monitored during a so-called oral glucose tolerance test, and the noise is correlated with a glucose reference measurement performed using the Abott FreestyleLibre 3 system during the same time period. Signal descriptors, such as signal-to-noise ratio (SNR), standard deviation (SD), etc., are calculated from the decomposed signal. ), perfusion index (PI), etc.

[0019] Clearly, tissue becomes more transparent as glucose levels, or more generally, the level of the substance being measured, increases. This is demonstrated by the decrease in DC levels during the same time period using the reflectance mode configuration. The observed behavior may suggest that the detected light has interacted with deeper tissue layers, and various physiological mechanisms, such as blood flow, muscle activity, etc., may add relatively more noise to the system. This interaction with deeper layers differs from that of the superficial layer, where there is little or no blood flow and the skin is composed of a rather passive layer of material.

[0020] It can be hypothesized that increasing glucose concentration opens an optical window for optical measurements toward deeper tissue layers. This hypothesis is supported by the fact that when glucose levels rise, the PPG modulation amplitude (AC, PI) essentially increases (e.g., 3.8 times here): this can only occur when the partial weight of the signal modulated by pulsating blood increases.

[0021] As a result of this observation, the inventors have proposed a novel method for determining the concentration of substances in samples containing liquids, particularly glucose from blood, where the refractive index of the liquid depends on the concentration of substances dissolved therein and the particle density in the liquid is substantially constant. The proposed method is based on extracting the noise spectrum from PPG data, and particularly separating the signal portion from the noise portion within the spectrum measured by PPG.

[0022] Although this method was primarily developed for glucose measurement, it is not limited to this and can be generalized to a method for determining substances dissolved in liquids containing particles, provided that the liquid is excited by pressure or volume modulation.

[0023] By obtaining the standard deviation from the noise spectrum and relating it to blood glucose or, more generally, the concentration of a substance in a liquid, the change in the substance level can be obtained, and if a baseline reference is known, the overall absolute value can also be obtained.

[0024] Several further results can be obtained from this method. For example, the ratio sigma can be used. 2 / DC is used to quantify the deviation from shot noise limiting performance and correlate this deviation with the concentration of substances in the blood. In some other aspects, the heart rate can also be calculated from the pure modulated signal and correlated with blood glucose concentration. Finally, the so-called modulation-to-noise ratio can be calculated, which quantifies the overmodulation of the signal due to heartbeat relative to system noise (including noise from the sample).

[0025] In some aspects, a method has been proposed for determining the concentration of substances in a sample comprising particles, particularly glucose in blood, wherein the refractive index of the liquid depends on the concentration of substances dissolved therein and the particle density in the liquid is substantially constant. In the proposed method, the liquid is modulated in its volume or in its density. If the liquid is blood, and the concentration of substances in the blood is to be measured, the heartbeat provides the necessary modulation.

[0026] To determine the concentration of a substance, a first signal is acquired during a first time period. The acquisition rate of the first acquired signal is set to be at least twice as large as the periodicity or frequency of the liquid modulation to satisfy the Nyquist criterion. For example, the acquisition rate can be 10, 20, or even 100 times larger. Typically, the acquisition rate is at least 40 times larger than the modulation frequency. In the case of blood with a heartbeat of approximately 1 to 2 Hz, the acquisition rate can be at least 10 Hz, and more particularly between 20 Hz and 200 Hz.

[0027] Similarly, the first time period is adjusted so that multiple periods of liquid volume or pressure modulation are adapted to the first time period. For example, the first time period can be 3 to 30 times longer than the liquid modulation time period. At a heart rate of 1 Hz, the acquisition time can be between 5 and 20 seconds.

[0028] This method ensures that noise is captured with sufficient resolution and a sufficient number of signal peaks during the first time period and the first signal acquisition.

[0029] In the second step, particularly in the frequency domain, at least a portion of the noise spectrum is extracted from the acquired signal. The frequency-dependent amplitude of the noise portion in the first acquired signal is estimated.

[0030] Regarding the term "part," it should be understood that the overall signal spectrum itself is band-limited due to real-world signal processing. It includes a noise portion, referred to as the AC portion and the DC portion of the actual signal. For the proposed method, a portion of the noise spectrum is obtained, estimated, or extracted from the overall signal spectrum (these terms are used synonymously in this application). Various methods can be employed for such extraction, some of which are presented herein. The extracted noise portion can be further band-limited and includes only noise reaching a specific frequency. This approach is adopted because systematic noise (which is the part of interest in the proposed method) attenuates at 1 / f, and noise at higher frequencies is uncorrelated with substance concentration.

[0031] In the final step, the substance concentration is estimated and determined based on an extracted portion of the noise spectrum and a reference value. Alternatively, changes in substance concentration can be determined based on an extracted portion of the noise spectrum and a reference value. In this regard, the reference value can be zero or any other benchmark value. For example, the reference value could correspond to a ground or normal level of substance concentration in the liquid.

[0032] The proposed method leverages the fact that, in some cases, substance concentration can be determined by deviating from a shot noise-limited system because the substance or its alteration introduces an additional term of systematic noise. This method may be beneficial in highly noisy systems or when the signal-to-noise ratio deteriorates significantly.

[0033] In some aspects, the reference value is based on a processed second signal. Therefore, the proposed method includes the following steps: acquiring a second signal during a second time period at an acquisition rate at least twice the periodicity of the liquid's volume or pressure modulation. The second time period can be after or before the first time period and has a certain time interval between measurements. Further measurements can be performed to obtain multiple such signals. It is preferable that the first and second time periods are the same to reduce possible residual effects, but subsequent method steps can compensate for different periodic measurements, for example, through normalization.

[0034] Specifically, in the frequency domain, a portion of the noise spectrum is extracted from the second acquired signal. The reference value is, in some aspects, based on the extracted portion of the noise spectrum of the second acquired signal. This allows for the acquisition of changes in substance concentration between two or more measurements. In some cases where substance concentration changes over time, the rate of change of substance concentration and other parameters can be obtained by evaluating multiple acquired signals according to the proposed method. Similarly, in some aspects, the reference value is given by a predetermined estimate or a value derived from that estimate.

[0035] In some aspects, the step of estimating at least a portion of the noise spectrum includes the following steps: identifying AC components in the spectrum of the acquired signal, which correspond to the fundamental frequency and its harmonics modulated by the volume or pressure of the liquid. Typically, the harmonics appear as multiples of the fundamental frequency. The identified AC components are then removed or separated from the spectrum, leaving the noise and DC components.

[0036] It should be noted that noise also exists at frequencies corresponding to the fundamental and harmonic frequencies modulated by the volume or pressure of the liquid. In essence, the noise component is submerged beneath the peak values ​​of the actual signal. Simply removing the peak values ​​at these frequencies would also remove or separate this noise component, but this is not ideal because it contains valuable information. Therefore, the inventors have proposed a method to estimate the noise component beneath the peak values ​​of the fundamental and harmonic frequencies.

[0037] Therefore, the process of removing the identified AC portion from the spectrum includes the following steps: estimating the noise portion of the spectrum at the fundamental frequency and its harmonics, which are modulated by the volume or pressure of the liquid, based on the values ​​of adjacent points in the spectrum. This step first defines the width of the noise portion to be estimated using the fundamental frequency peak width or a given mask. The estimation itself uses the noise directly adjacent to the corresponding peak. This is based on the assumption that the noise below the peak is constant. As an alternative method, the noise portion below the masked portion of the spectrum can be estimated based on points at the boundaries of the masked portion of the spectrum.

[0038] Therefore, the gaps in the noise spectrum caused by the removal or separation of the AC section are filled with the average value of adjacent points.

[0039] In some other aspects, the signal processing used to identify the AC portion of the spectrum of the first and second acquired signals can comprise several different tasks. These can be used individually or in combination with other steps to distinguish peak signals from noise components.

[0040] In some applications, smoothing functions are applied to the spectra of the first and second acquired signals. Smoothing reduces noise, so it should not be too strong, but it also helps to reduce outliers. Smoothing functions can be applied before obtaining the spectrum to smooth out any unevenness that may exist in the harmonics.

[0041] In some respects, particularly by raising the absolute value of the spectrum to the power of M (where M is an integer greater than 2), at least the fundamental frequency peak can be enhanced. This is especially useful when the data is noisier and the peak in the spectrum only slightly outshines the adjacent noise. By applying this function, the peak is significantly enhanced relative to the noise, which also simplifies subsequent processing steps.

[0042] In some aspects, a mask is defined, which has a width based on a predetermined amplitude level of at least the fundamental frequency peak. For this purpose, a reference level can be set to define the mask boundaries, for example, at 10% or 5% of the peak level. Other levels are also possible, such as 3%, 2%, or even 1% and below. The reference level used to define the noise can depend on the overall signal level (i.e., a portion of it) and also on the enhancement level.

[0043] In some aspects, the dynamic peak width is determined, for example, by f max -f min The frequency is given by the corresponding reference level at the rising and falling edges of the peak values. In some respects, a static offset can be added to the dynamic width, but the advantage is that this offset only needs to be determined once for all measurements.

[0044] In some aspects, the centroid of at least the fundamental frequency peak can optionally be determined by using a width based on a predetermined amplitude level of at least the fundamental frequency peak. This centroid can be determined by summing the product of the frequency and the FFT amplitude over a previously determined width of the fundamental frequency. The result is then divided by the sum of the amplitudes. It should be noted that the centroid of the fundamental frequency peak does not necessarily correspond to the maximum value; this only applies to symmetrical peaks, which are rare. The centroid of the fundamental frequency also provides information for identifying the correct frequencies of harmonics.

[0045] A mask of a defined width is applied over a predetermined number of possible peaks, and more particularly over the corresponding centroid of the fundamental frequency, and then repeated at the harmonic frequencies. The peak values ​​of the harmonics are calculated based on the fundamental frequency within the spectrum, which has been determined as described above. Thus, the mask covers a predetermined number of peaks corresponding to the fundamental frequency modulated by the volume or pressure of the liquid and its harmonics, thereby identifying the AC component.

[0046] As mentioned above, the AC portion determined by this method can then be separated from the spectrum. Not all harmonics are visible or contribute to the spectrum due to noise levels. Therefore, in most cases, it may be sufficient to consider only the fundamental frequency and its harmonics up to the seventh order, and especially the fifth and fourth orders. However, this is not a limitation, as some applications may require higher orders, such as the 10th or higher harmonics. This may depend on where the measurement is taken (e.g., finger and wrist) and the wavelength. Depending on the location, a relatively large number of harmonics can be resolved.

[0047] In some cases, the first and second signals are preprocessed. For example, the first and second signals can be filtered to remove transients, especially those with frequencies below 2 Hz, and especially below 1 Hz, and especially below 0.5 Hz. Heart rate at or below 0.5 Hz is quite rare. Respiration typically occurs in the 0.2 Hz range (5 seconds per respiratory cycle / 12 breaths per minute) and falls within the <0.5 Hz range. These transients can be generated by respiration, for example, in the case of measuring blood. In some cases, preprocessing is performed by applying a moving average filter with a window width approximately equal to one heartbeat, which captures slow transients. The transients are then removed and the DC portion is re-added to generate a PPG signal that is fairly flat but still retains the AC modulation due to the heartbeat.

[0048] In some aspects, background subtraction has been performed on the measured signals to remove dark counts and ambient components. The first and second signals can be converted into digital signals before further processing. The aforementioned preprocessing can be performed in the analog domain (e.g., low-pass filtering and background subtraction), while the Fourier transform used to extract the noise component is performed in the digital domain. In some aspects, the first and second acquired signals are processed using a Fast Fourier Transform.

[0049] Some aspects involve the calculation and determination of substance concentration or changes in substance concentration. In some aspects, the standard deviation of an estimated portion of the noise spectrum is obtained. The standard deviation σ is given by the following formula: , where N is the number of samples within the frequency range of interest.

[0050] Let M be the number of measurement points for each data trajectory, for example, 2048, measured at a rate of 200 Hz, with a measurement time of approximately 10 seconds (approximately 10 heartbeats). This results in a frequency scale of (-M / 2, M / 2) / 2048 × 200 Hz = (-1 / 2, 1 / 2). 200Hz = (-100Hz, 100Hz), with a frequency step size of 2048.

[0051] In this example, the frequency step size would then be 200Hz / 2048, approximately 0.1 Hz. If you want to refine the frequency step size / resolution, simply increasing the sampling rate is not enough; you need to measure over a longer time period, such as measuring 4096 points at 200Hz.

[0052] If N is set to be equal to M, we obtain the standard deviation using a common formula in the time domain; this is Passevar's theorem. This relation can be used to evaluate the standard deviation over the entire frequency range, or only over a portion of it. It determines the starting point (i=2, or another number) and the ending point: for example, at 0.5 Hz and 10 Hz, in the example (200 Hz, 2048 points) N = approximately 102, and the starting point is at approximately 5.

[0053] Standard deviation is related to substance concentration; therefore, given a reference value, the substance concentration or its change can be directly obtained simply by evaluating the standard deviation. In some respects, the signal-to-noise ratio of an estimated portion of the noise spectrum to the reference value, or its reciprocal, can be used, as it is also related to substance concentration.

[0054] In some respects, the concentration of a substance or a change in its concentration can be obtained by evaluating the quantification of the deviation from the shot noise limiting system as defined by the ratio of the standard deviation of the estimated portion of the noise spectrum. As previously mentioned, due to correlation, there exists a systematic noise component that deviates from the classical conventional shot noise limiting system. Several signal descriptors suitable for determining the concentration of a substance or a change in its concentration based on the proposed principles are presented in this paper.

[0055] In other aspects, the proposed method also includes calculating the modulation-to-noise ratio (MNR), which is given by the AC component relative to the estimated noise. The MNR quantifies the degree of signal modulation due to the heartbeat relative to the system noise.

[0056] The proposed method can be applied, in particular, to the determination of substances in the blood, such as glucose or O2. In this case, when the fluid is blood, modulation is caused by the heartbeat, resulting in changes in blood pressure and volume. Obtaining such a signal refers to PPG measurement. In some aspects, heart rate is calculated based on the AC portion identified in the spectrum. A correlation between heart rate and blood glucose concentration has also been found, meaning that heart rate appears to change as blood glucose concentration increases or decreases.

[0057] Some aspects relate to a detector device. The detector device can be implemented in a single housing, such as a watch or medical instrument, and can also include distributed components. In some aspects, a detector device is proposed for determining the concentration of a substance in a sample comprising a liquid containing particles, particularly glucose in blood, wherein the refractive index of the liquid depends on the concentration of the substance dissolved therein and the particle density in the liquid is substantially constant. The detector device includes at least one light source and at least one detection element, wherein the at least one detection element is optically separated from the at least one light source. The at least one light source is configured to emit light through an exit window onto the sample containing the liquid with the substance.

[0058] The detection element is configured to detect the light component corresponding to the emitted light scattered through the particulate liquid. The detector device also includes control circuitry coupled to at least one light source and at least one detection element. The detector device is configured to perform the method steps described above.

[0059] In some other aspects, the detector device may include evaluation circuitry for extracting at least a portion of the noise spectrum from the first acquired signal, particularly in the frequency domain, and for obtaining a substance concentration or a change in substance concentration from the extracted portion of the noise spectrum. The evaluation circuitry may be implemented in a single housing along with other components, or it may be arranged separately and communicate using a wired or wireless interface. For example, the evaluation circuitry may be a computer program executing on a mobile device, computer device, or medical device.

[0060] In some aspects, the control circuit is configured to control at least one light source to emit light signals during a first time period at multiple different consecutive times and to obtain signals from at least one detection element. The control circuit can be controlled and triggered by an evaluation circuit.

[0061] Some aspects relate to the implementation of the detector device. The detector device may include multiple photodetectors arranged in a ring or quadrilateral shape, and optionally, particularly, arranged at different distances around at least one light source, particularly centrally. In other aspects, the at least one light source includes multiple optoelectronic devices arranged at different distances from at least one detection element; and optionally, includes a ring or quadrilateral shape, optionally arranged around at least one detection element, particularly centrally.

[0062] It may be suitable if at least one light source is configured to emit light of a different wavelength. This can compensate for different skin types and other parameters. In some other aspects, at least one detection component includes a filter that includes low transmittance in a spectrum different from the spectrum emitted by the at least one light source.

[0063] Although the optoelectronic device presented herein is described only in terms of its functionality in determining substance concentration, it should be noted that various implementations are possible. In this respect, the evaluation unit does not need to be implemented within the housing containing the transmitter and detector, but can be located separately from them. In some aspects, the evaluation unit is implemented in a separate device separated from the housing itself. Communication between the transmitter and detector, and the evaluation unit as described above, is facilitated, for example, via wireless communication. This would allow, for example, a master-slave configuration, where the evaluation unit requests measurements periodically. Furthermore, the evaluation unit can be implemented primarily in software, for example, as an application running on a mobile device, while the rest of the optoelectronic device is implemented in a separate housing.

[0064] In some respects, the housing (or more generally, the optoelectronic device) is implemented as a ring, earplug, watch, or any other wearable that can adapt to the user's daily environment and be continuously worn. These rings, earplugs, watches, or other wearables communicate with an evaluation unit or a mobile device or other device that implements the evaluation unit. The ring, earplug, watch, or other wearable can cover a large area of ​​the user's skin, such as wrapping around a finger, clipping to an ear, or inserting into an ear, and can be used to take measurements at one, opposite, or multiple locations, etc.; measurements can also be envisioned on opposite sides of the wrist, such as the dorsal (top) and palmar sides near the buckle of a wristband. They can contain multiple transmitters and detectors located at various locations, enabling them to take measurements not only at a single point but also simultaneously or sequentially at multiple points. As a result, skin irregularities or other problems can be overcome, and overall measurement quality can be improved.

[0065] Besides wearable and handheld devices, other applications are also possible. For example, optoelectronic devices can be implemented in medical devices or laboratory instruments, such as for testing and measurement purposes. These devices can also be stationary or mobile.

[0066] Some further aspects involve mobile displays in which detectors are directly implemented. In such applications, display LEDs (e.g., for red and green) can be used as emitters according to the proposed principles. A finger is placed directly on the display surface and then illuminated through the display to obtain a first and / or second signal. Similarly, the proposed principles can be implemented in VR or AR glasses and devices. Another application involves safety issues, such as during certain strenuous or dangerous work, while driving a motor vehicle, etc. Such optoelectronic devices can be implemented in cars, for example on a steering wheel, according to the proposed principles, to obtain a perfusion index and thus the noise spectrum during driving. This makes it possible, for example, to warn the driver of potential health threats while driving.

[0067] Implementing a remote PPG sensor is also possible and has been validated: in such systems, a light source remote from the sample is imaged using an optical lens. Similarly, the detector is positioned in a plane conjugate to the sample in the imaging configuration. The size of the illumination spot is controlled by the size of the light source, and the NA (aperture detector) can be controlled by the aperture in the collimation space, i.e., by the size of the net aperture of the lens. The same considerations apply to the detection side, where the size of the collection spot is defined by the area of ​​the detector. With this setup, enhanced sensitivity can be achieved simply by having the illumination and collection paths intersect at a certain distance above the sample, through negative illumination and negative detection angles. Attached Figure Description

[0068] In contrast to the various embodiments and examples described in detail with reference to the accompanying drawings, other aspects and embodiments based on the proposed principles will become apparent, as illustrated in the drawings.

[0069] Figure 1 A detector device based on some aspects of the proposed principles is shown; Figure 2 It shows that it can be done Figure 1 Exemplary PPG measurements obtained by the detector device; Figure 3 The reference curve shows the change of glucose concentration over time and the corresponding PI measurement results obtained using light in the near-infrared spectrum, with two points selected to illustrate some aspects of the proposed principle. Figure 4 Showing from Figure 3 The two PPG signals at points #1 and #30; Figure 5These are the corresponding spectra from the previous two PPG measurements; Figure 6 A detailed view of the enhanced spectrum and fundamental frequency of data point #1 is shown; Figure 7 An exemplary mask generated from information obtained from the fundamental wave shown in the previous figure is illustrated, based on some aspects of the proposed principles; Figure 8 The application of the generated mask to the enhanced spectrum of data point #30 is shown respectively; Figure 9 The spectrum of point #30 is shown according to some aspects of the proposed principle, where the masked portions corresponding to the fundamental and harmonic frequencies are separated from the noise; Figure 10 The estimated noise below the corresponding peak values ​​of the fundamental and harmonics separated from the noise is shown based on some aspects of the proposed principle; Figure 11 The fundamental and harmonic frequencies are shown after subtracting the estimated noise, based on some aspects of the proposed principle. Figure 12 The spectra of points #1 and #30 are shown respectively; Figure 13 The extracted noise spectra at points #1 and #30 are shown based on some aspects of the proposed principle; Figures 14 to 21 Multiple descriptors based on the proposed principles and their respective correlations with changes in glucose concentration are shown. Detailed Implementation

[0070] The following embodiments and examples disclose various aspects and combinations thereof based on the proposed principles. The embodiments and examples are not always to scale. Similarly, different elements may be shown enlarged or reduced in size to emphasize aspects. It goes without saying that the aspects of the embodiments and examples shown in the figures can be combined with each other without contradicting the principles of the invention. Some aspects exhibit regular structures or forms. It should be noted that slight differences and deviations from the ideal form may occur in practice, but this does not contradict the concept of the invention.

[0071] Furthermore, the figures and aspects are not necessarily shown at the correct dimensions, and the proportions between the elements are not necessarily substantially accurate. Some aspects are highlighted by magnification. However, terms such as "above," "above," "below," "under," "larger," and "smaller" are correctly represented in relation to the elements in the figures. Therefore, such relationships between elements can be inferred from the figures.

[0072] Figure 1A detector device based on some aspects of the proposed principles is illustrated. In this embodiment, the detector device facilitates glucose measurement in the blood vessels of a human or animal body. However, it should be understood that the invention and the proposed method are not limited to such measurements. Instead, for example, the concentration of a substance in a liquid flowing through an artificial tube can be measured using an artificial tube instead of human tissue. The various elements of such a detector device are similar to... Figure 1 The apparatus shown in the figure. This enables the proposed method to be used for a variety of measurements in different applications.

[0073] This device is used for various measurements, including, for example, PPG or SpO2 measurements. In this regard, the proposed method, as explained herein, can be implemented in existing hardware via software, provided the detector device is suitable for providing a PPG signal including a noise component. In this embodiment, the detector device 1 includes a housing 11 that includes at least one light source 12 configured to emit light pulses of at least one wavelength in the visible or IR spectrum.

[0074] Light source 12 may also include multiple light sources (not shown here) configured to emit light of different wavelengths or located at different positions. This allows for compensation for skin irregularities and also provides adjustment for the signal-to-noise ratio, as light penetration to deeper layers is strongly wavelength-dependent. Typical wavelengths suitable for PPG measurements and measurements used to determine glucose concentration include light in the green and red visible spectra as well as the near-infrared spectrum.

[0075] The housing 11 also includes a detector device 13, which is arranged separately from the light source 12. A light-locking element 14 is arranged between the light source 12 and the photodetector. In this example, the photodetector 13 includes multiple detector regions 13.1, 13.2, etc. The detector regions are located at different distances from at least one light source. The various distances between the light source 12 and the detector regions 13.1, 13.2 can compensate for skin irregularities and also provide distance-dependent signals. These signals can also be used to determine substance concentration, as the total amount of scattering depends on the distance.

[0076] In this invention, due to the increase in systematic noise caused by the scattering of emitted light within the user's skin, multiple detector regions 13.1, 13.2 at different distances from the light source 12 enable the determination of an optimal distance, wherein the ratio of systematic noise (used in the proposed method) to other noise components is maximized.

[0077] Finally, the detector device 1 also includes an evaluation and control device 10. The evaluation device 10 may be placed within the housing 11 of the device or positioned separately. For example, the housing 11 may be part of a smartwatch or a wearable device that communicates with another device via a wireless interface. Other devices may include the evaluation and control device 10. The evaluation and control device 10 may be implemented as hardware, software, or a combination of both.

[0078] The evaluation and control device 10 controls the measurement and can be considered as an integrated data acquisition system that can provide control signals to a current driver for at least one light source 12, an analog-to-digital converter for sampling the current from detector device 13 and regions 13.1, 13.2, and a clock for triggering and timing the measurement.

[0079] During measurement, the user places their finger, which has tissue 30, on surface 20, which is part of glass interface 21 or another transparent interface 21. Interface 21 is transparent to light emitted by a light source, but may be opaque or have reduced transparency for another wavelength. This reduces dark current in photodetector 13 and its regions 13.1, 13.2.

[0080] Light emitted by light source 11 travels through tissue 30 along various optical paths 32 within tissue 30, where interactions occur with various blood-carrying and non-blood-carrying layers. Due to the resulting periodic blood flow, the heartbeat modulates the portion of blood within the measurement volume, thereby encoding a modulation on the light flux re-emitted by the tissue and detectable.

[0081] The measurement modulation corresponding to the heartbeat contains various information that can be used for determination and is commonly referred to as a PPG measurement. For the purposes of this application, the three components will be explained in more detail.

[0082] Figure 2 This PPG signal is shown. The PPG signal includes a signal level or DC component (approximately 5.2 × 10⁻⁶ in the example shown). 4 The heartbeats are represented by (1) the main modulation AC of the signal due to the heartbeat (which is usually the object of direct concern in various applications) and (3) the noise component (most easily identified in the example shown by significant point-to-point variability). The heartbeats are given by the maximum value in the signal and are in the range of approximately 60 beats / minute (1 Hz). The perfusion index, or PI, is defined by the ratio of AC to DC, approximately 0.1 × 10⁻⁶. 4 The magnitude of the AC portion of the count is divided by the DC portion. In the given example, PI is approximately 1.9.

[0083] The measurements used for PPG measurements and those performed by the proposed method are typically acquired over multiple heartbeats, for example, within 5 to 20 seconds, and the acquisition rate is generally greater than the heart rate. Because the proposed method requires a deep understanding of the noise applied to the actual signal, it is suitable for acquiring PPG measurements at a rate at least 20 times higher than the actual heartbeat rate. Since the heartbeat itself varies between approximately 60 beats / minute and approximately 180 beats / minute, a data acquisition rate between 50 Hz and 200 Hz is suitable. Figure 2 In the example depicted, a sampling rate of 100 Hz was used, and each trace had 1024 data points.

[0084] Below, a method is shown for decomposing one or more PPG measurements into DC, AC, and noise components and using the noise component to determine glucose concentration. Glucose determination using AC and DC components has been demonstrated and is subject to other applications filed by the applicant. While several different methods exist for extracting noise from signals in the time and frequency domains, the proposed method applies a frequency domain approach.

[0085] Figure 3 A typical oral glucose tolerance test is shown, where the reference data, referred to as the reference, is taken from the Abbott Freestyle Libre 3 system. This constitutes the reference measurement during the test. Food intake occurs at approximately 10:20, indicated by the blue vertical bar and the absence of a data point. Shortly thereafter, optical and reference measurements are performed every minute. The optical measurement lasts 10 seconds, a period long enough to obtain a suitable PPG measurement and short enough that the change in glucose concentration during a single measurement is not excessive.

[0086] After consuming sugary foods, glucose concentration rises over time, peaking about 40 minutes after eating and then declining over the next 40 minutes. Figure 3 The optical measurement (represented by IR PI) is the perfusion index of IR light at 940 nm. The response of PI to blood glucose concentration is subject to another separate application filed by the applicant.

[0087] For the purpose of noise extraction, PPG traces from indicator points #1 and #30 will be used in the exemplary embodiment. Of course, the proposed method can be performed for each data point, i.e., each PPG measurement. The exemplary points correspond to low and high blood glucose concentrations, respectively.

[0088] Figure 3The actual PPG traces of data points #1 and #30, obtained by detecting light scattered by human tissue, are shown. The measured signals were preprocessed using a low-pass filter to remove slow transients, such as those caused by respiration. Transient removal is suitable for suppressing endpoint discontinuities, which typically increase the 1 / f component in the frequency domain. Similarly, dark current is at least partially removed. As can be seen from both traces, rising glucose concentrations lead to an increase in the AC component and a decrease in the DC component. Therefore, point #1, with its low glucose concentration, comprises approximately 0.6 × 10⁻⁶ PPG. 4 It has a higher DC component and a lower AC component, which is also quite noisy. The trace at point #30 includes approximately 4.8 × 10⁻⁶. 4 It has a lower DC value and a higher AC component. The noise can resemble other noises, at least from the perspective of visually perceptible noise.

[0089] After signal preprocessing, a Fourier transform is performed and the signal is normalized to the corresponding DC value (zero frequency) to provide a relative comparison of the modulation amplitude. Figure 5 As can be seen, the fundamental frequency of approximately 1.3 Hz from the heartbeat is clearly visible in both traces. However, the amplitude at point #1 decreases, and the harmonics are almost invisible compared to the trace at point #30. The amplitude of the fundamental frequency of approximately 1.2 Hz at point #30 is about three times greater than that at point #1, which corresponds to a lower glucose concentration.

[0090] like Figure 5 As shown, the noise floor that can be distinguished between harmonics is also higher as glucose concentration increases. This observation is new, and it could provide additional means for non-invasively detecting blood glucose via optical measurements and, with proper calibration, determining changes and even their absolute values.

[0091] The following figures illustrate a method for extracting noise levels from a spectrum and correlating those noise levels with glucose levels to determine concentrations based on changes in noise over time, and more particularly in PPG measurements.

[0092] In the first step, further preprocessing is performed to enhance the fundamental frequency and the peak values ​​of harmonics relative to noise. This enhancement is achieved by raising the absolute value of the FFT to the power of M, where M can be an integer such as 2, 3, 4, etc. Enhancement is necessary when the actual signal from the heartbeat is relatively small or the noise level is high, ensuring that the harmonics only slightly outshine adjacent noise. Furthermore, a smoothing function is applied, in the exemplary embodiment a 3-point smoothing function, where a moving average is applied to three points of the PPG measurement. The result of this additive preprocessing is... Figure 6 The fundamental frequency is visible at point #1, where M=6 and a 3-point average is used.

[0093] The peak amplitude is significantly greater than the noise level. In the first aspect, the frequency f is determined. min and f max The amplitude reaches a specific level, for example, 5% of the overall maximum value in this case. Other peaks can also be selected and may depend on the enhancement factor. Level detection determines the dynamic peak width as f. max -f min To obtain the actual peak width, a static offset can be added to the dynamic width, but the advantage is that this offset only needs to be determined once for all measurements. Furthermore, the fundamental peak value is asymmetrical, so the centroid must be determined using the dynamic width in the next step, but at the intermediate frequency (f... max -f min The offset is the difference between ) / 2 and the center of gravity.

[0094] This can be done by summing the product of the frequency and the FFT amplitude over the dynamic width of the previously determined fundamental frequency. The result is then divided by the sum of the amplitudes. Figure 6 and Figure 7 In the illustrated example, the center pixel is at position 13.93 (an arbitrary value), which roughly corresponds to a frequency of 1.27 Hz. Although the centroid roughly coincides with the maximum value in this example, it doesn't have to, and it depends on the amplitude distribution. The final mask of the fundamental peak boundary is... Figure 7 As shown, it was applied to the original FFT of point #1, where the baseline level of the peak was approximately 5%.

[0095] A mask, with its width and center frequency already determined, is then applied to the spectrum. More specifically, the total number of masks to be applied to the spectrum can be adjusted, typically between 3 and 10. The reason for the limited number of masks used to mask harmonics and the fundamental frequency is that harmonics decay very quickly and are soon completely submerged in noise. Therefore, in most cases, it is unnecessary to request more than 10 harmonics, as they will not contribute any further.

[0096] For this process, the mask is placed with the corresponding width of the fundamental peak value obtained in the previous step and at a center frequency corresponding to a multiple of the fundamental. Additional offsets can be applied if necessary. This process is repeated for all higher harmonics until the maximum number of harmonics to be resolved and analyzed is reached.

[0097] Figure 8 The mask obtained for the fundamental frequency at point #30, where up to the ninth harmonic is applied, is shown. Note that the mask width and its centroid are determined separately over time for each measurement point. This is necessary because the heart rate may change over the overall time period.

[0098] Now, the mask is multiplied with the spectrum of point #30, thus separating the fundamental frequency and peak value (plus the noise component below the peak value) from the noise between the individual masks. The result is... Figure 9 As shown, curve K1 represents the peak values ​​of the fundamental frequency and harmonics up to the ninth harmonic. Curve K2 provides the noise portion of the spectrum between masks.

[0099] Figure 10 The subsequent steps for estimating the noise beneath the masking peaks of the fundamental and harmonic frequencies are shown. For this purpose, it is assumed that the noise is frequency-independent and constant over the masking range, i.e., for each mask at f... min to f max The frequency dependence can also be assumed to be constant. Since noise typically decays with increasing frequency, a slight frequency dependence can also be assumed.

[0100] To estimate the noise below the peak, the gaps in the noise spectrum caused by the mask are filled with the average value of adjacent points, corresponding to the limits of the mask at lower and higher frequencies. The results are as follows: Figure 10 As shown, the noise level between the peak values ​​of the fundamental and harmonic frequencies has been removed for viewing purposes.

[0101] Since the spectrum includes complex numbers, an additional step can be taken by fitting the phase of the noise spectrum in the peak regions. The goal is to minimize the difference between the complex data amplitude and the complex noise amplitude at each frequency location, using the noise amplitude determined above without modification. This phase fitting step is an approximation, obtaining a conservative estimate by assuming that the noise will always reduce the modulation amplitude. However, from a statistical perspective, noise can also artificially enhance the modulation, but since the actual phase difference between modulation and noise is unknown, the conservative method has been observed to provide satisfactory results.

[0102] When comparing Figure 9 When considering curves K1 and K2 in the spectrum, the advantages of the proposed method become apparent. It is unnecessary to know the number of present or identifiable harmonics in the spectrum, as higher and weaker harmonics become indistinguishable from noise and therefore have no effect on the modulation spectrum. This allows specifying, as previously stated, the number of harmonics to be considered for analysis, and knowing that only those harmonics that stand out above the noise floor will have an impact. This resolves a previously challenging problem.

[0103] Figure 11 The peak levels of the fundamental and harmonic frequencies are shown after subtracting the estimated noise level. In this example, only the first two harmonics contribute significantly to the modulation spectrum, which is consistent with the methods described above and their advantages.

[0104] Figure 12 The overall spectra of points #1 and #30 are shown respectively. Figure 13These are the noise spectra corresponding to the two points, where the estimated noise fills the gaps between the removed peaks. Therefore, the AC part in Figure 13 The frequency spectrum is separated, leaving only noise and DC (at frequency 0).

[0105] For each PPG measurement, a spectrum is obtained, and then the spectrum is further processed. Multiple descriptors are available for processing the noise spectrum, as well as combinations of one or more other components with the noise spectrum. All these descriptors show a correlation with glucose concentration, and therefore they are affected by the substance concentration or its changes. A useful description is the standard deviation σ, given by the following formula: Where DC is the DC part and N is Another description is given through absolute signal modulation (i.e., without...). Figure 11 (The noise shown below in the spectrum). The modulation AC is given by the following equation:

[0106] The perfusion index PI is known and is given by the ratio of AC / DC.

[0107] It has been observed that, in the absence of any effect of substance concentration on noise, the system behaves similarly to a shot noise-limited system. However, a bias has been observed, which itself is related to the concentration of glucose or, more generally, the substance concentration. The shot noise-limited system is given by the sigma σ equal to the square root of the DC part. Therefore, the signal-to-noise ratio (SNR) can be expressed as...

[0108] Finally, the modulation-to-noise ratio (MNR) is given by the following formula:

[0109] the following Figures 14 to 21 The correlation between various descriptions and glucose concentration or its changes is shown.

[0110] The obtained data were not smoothed by averaging and may therefore appear noisy. However, the correlations are clearly visible, and further averaging of the individual points in the PPG measurements (i.e., moving averages over 3 to 5 individual measurements) would greatly smooth the curve. Furthermore, a reference curve is also shown in the figure below and has been shifted (the original curve is shown for reference). Optical measurements appear to be more sensitive. Therefore, it is affected approximately 5 to 10 minutes earlier than the reference curve. This could be explained by the delay until the rise in glucose in the interstitial fluid, or by other reasons, such as changes in blood concentration occurring before those in the interstitial fluid, or by other changing parameters in the tissue.

[0111] To improve readability and possibly refer to existing relevant information, Figure 14 The perfusion index PI for 940 nm light and a reference measurement are shown. The reference curve is also used in the following figures. For all measurements during the oral glucose tolerance test (OGTT), the perfusion index PI was calculated from the pure AC spectrum using the ratios described above, as shown. Figure 11 As shown, for reference measurements varying between 82 mg / dL and 155 mg / dL (corresponding to an approximately 89% increase in peak value relative to minimum), the PI changed from approximately 0.0065 to approximately 0.031, representing an increase of nearly 3.8-fold over the same time period. This demonstrates the sensitivity of the method, including both PPG measurements and the analytical algorithm proposed in this paper.

[0112] Figure 15 The DC levels (1 / DC) during the same OGTT are shown. As discussed in a previously unpublished application by the applicant, in a reflective PPG configuration, DC levels decrease as glucose concentration increases, and vice versa. The peak values ​​are approximately 20% higher than those observed during normal glucose levels at the start and end of the OGTT.

[0113] Now for reference Figure 16 The figure illustrates the standard deviation σ calculated above. Besides its inherent variability, it also shows a strong correlation with the glucose reference measurement. This observation is unexpected because the DC level decreases with increasing glucose. The decrease in DC implies that σ also decreases. While this holds true for shot noise-limited systems, this behavior appears to deviate from the shot noise limit as glucose concentration increases.

[0114] This behavior is observed Figure 17 It becomes particularly obvious at times. Figure 17 The curve is shown plotting the standard deviation σ against the square root of the DC portion and the DC portion itself. As expected, the square root of DC decreases as DC decreases, but the standard deviation σ increases with increasing glucose concentration, and then decreases again. It can be concluded that system noise increases as glucose levels rise. If this finding is corroborated with a deeper understanding of light-tissue interactions, it could provide further insights into PPG measurements, and especially into optical non-invasive blood glucose measurements.

[0115] If the standard deviation σ increases while the DC portion decreases, then the SNR -1 =σ / DC should also be correlated with blood glucose concentration and its changes. This is in Figure 18 As shown in the figure. In other words, the signal-to-noise ratio, or the reciprocal of the signal-to-noise ratio, as defined herein, can be a measure of the concentration of a substance according to the proposed principle.

[0116] Figure 19 σ is shown2 The behavior of / DC serves as a measure of how close the data is to the shot noise limit performance. For a pure shot noise limit system, the noise level should produce a value of 1, which is equal to the shot noise or sqrt(DC). However, even at low glucose concentrations, such as at the start of an OGTT, the overall system noise exceeds the shot noise limit (which is approximately 1.25). Intake and increases in glucose concentration also show a further deviation from the shot noise performance, averaging approximately 2.25 at the highest glucose concentration levels.

[0117] Figure 20 Another description is shown, namely the absolute modulation AC relative to the standard deviation σ. This descriptor is also known as the modulation-to-noise ratio (MNR). This descriptor can be used as a quantitative measure and as an indicator of the performance of the proposed method in standard vital sign monitoring applications, such as heart rate monitoring or SpO2 measurement. The modulation-to-noise ratio quantifies the dominance of the modulation caused by the heartbeat relative to the noise in the system. The higher the overmodulation, the easier it is to perform measurements and extract one or more pieces of information needed for a given application.

[0118] at last, Figure 21 By extracting previously Figure 11 The pure modulation spectrum is shown, and the individual harmonics are summed using the algorithm to illustrate the heart rate during the oral glucose tolerance test. Overall, twelve channels were available for measurement in the system, and the results were superimposed. Despite some significant point-to-point variability, the resting heart rate increased slightly during the oral glucose tolerance test, indicating a correlation with glucose concentration. Similar observations were made in SpO2 measurements, which also correlated with glucose concentration.

[0119] Overall, the observation that noise is affected by substance concentration can support other applications and measurements, including SpO2 measurement. The proposed method is a useful complement to further enhancing non-invasive blood glucose measurement in any case. One benefit of this method is that it can be easily added to and complements existing measurements as well as other current PPG-based applications.

[0120] This method extracts noise from PPG measurements and uses it to determine the concentration of substances in a liquid. This is based on the fact that the underlying system of liquid, particles, and substances, as well as the light source and optical detector, deviates from the shot noise limiting system. This deviation is caused by systematic noise introduced into the system.

[0121] While a comprehensive theory explaining the detailed causes and origins of this observation may not yet exist, it is reasonable to hypothesize that the liquid becomes more transparent due to the reduced difference in refractive index. This is demonstrated by the decrease in DC levels during the same time period using the reflection mode configuration. Therefore, light from deeper layers, which has interacted more with the tissue and accumulated more noise, can be detected. This additional accumulated noise could originate from various physiological mechanisms, such as blood flow, neural activity, and muscle activity. This hypothesis is supported by the fact that the PPG modulation amplitude (AC, PI) substantially increases with increasing glucose levels. This increase can be explained by an increased weighting of the portion of the signal modulated by pulsating blood.

[0122] List of reference numerals

[0123] 1. Detector device

[0124] 10. Control and Evaluation Circuits

[0125] 11. Shell

[0126] 12 light sources, LED

[0127] 13 Detectors

[0128] Detector regions 13.1 and 13.2

[0129] 14 Optical Barrier

[0130] 21 Glass Interface

[0131] 20 Surface

[0132] 32 optical paths

[0133] 30 organizations

[0134] 32. Optical path.

Claims

1. A method for determining the concentration of a substance in a sample comprising a liquid containing particles, particularly glucose in blood, wherein, The refractive index of the liquid depends on the concentration of the substance dissolved in the liquid and the density of the particles in the liquid is substantially constant, wherein the volume or pressure of the liquid is modulated, comprising the following steps: - A first signal is acquired during the first time period at a sampling rate at least twice as large as the periodicity of the volume or pressure modulation of the liquid. - In particular, in the frequency domain, at least a portion of the noise spectrum is extracted from the first acquired signal; - Obtain at least one of the following: ○ The concentration of the substance derived from the extracted portion of the noise spectrum and the reference value; and ○ Changes in the concentration of the substance derived from the extracted portion of the noise spectrum and the reference value.

2. The method according to claim 1, further comprising: - After obtaining the first signal, a second signal is obtained during the second time period at a sampling rate at least twice as large as the periodicity of the volume or pressure modulation of the liquid; - In particular, in the frequency domain, at least a portion of the noise spectrum is extracted from the second acquired signal; Optionally, the reference value is based on the extracted portion of the noise spectrum of the second obtained signal.

3. The method according to any one of the preceding claims, wherein, The reference value is given by a predetermined extraction portion, or by a value derived from such extraction portion.

4. The method according to any one of the preceding claims, wherein, The step of extracting at least a portion of the noise spectrum includes the following steps: - Identify the AC portion of the spectrum of the acquired signal, the AC portion corresponding to the fundamental frequency and harmonics modulated by the volume or pressure of the liquid; - Remove the identified AC portion from the spectrum to obtain the noise portion.

5. The method according to claim 4, wherein, The step of identifying the AC portion of the spectrum of the first and second acquired signals includes at least one of the following steps: - Apply a smoothing function to the spectra of the first and second acquired signals; - In particular, by raising the absolute value of the spectrum to the power of M, at least the fundamental frequency peak is enhanced, where M is an integer greater than 2; - Determine a mask having a width based on a predetermined amplitude level at least of the fundamental frequency peak; - Optionally, the center of at least the fundamental frequency peak can be determined by using the width based on a predetermined amplitude level of at least the fundamental frequency peak; - The AC portion is identified by applying a mask, in particular a determined mask, to a predetermined number of peaks corresponding to the fundamental frequency and harmonics modulated by the volume or pressure of the liquid.

6. The method according to claim 4, wherein, The step of removing the identified AC portion from the spectrum includes the following steps: - Based on the values ​​of adjacent points in the spectrum, extract the noise portion at the fundamental frequency and harmonics modulated by the volume or pressure of the liquid within the spectrum; or - Based on points at the boundaries of the masked portion of the spectrum, extract the noise portion of the masked portion of the spectrum.

7. The method according to claim 4, wherein, The AC portion corresponds to the fundamental frequency and up to the seventh harmonic, and especially up to the fifth harmonic and especially up to the fourth harmonic.

8. The method according to any one of the preceding claims, wherein, The steps of obtaining the first signal and / or the second signal include at least one of the following: - Low-pass filtering is applied to the first acquired signal and / or the second acquired signal to remove transients, especially those with frequencies below 2 Hz, especially below 1 Hz, and especially below 0.5 Hz; - Remove the dark count from the first obtained signal and / or the second obtained signal; as well as - Convert the first acquired signal and / or the second acquired signal into a digital signal.

9. The method according to any one of the preceding claims, wherein, The step of extracting at least a portion of the noise spectrum from the first and / or second includes the following steps: - Perform a Fourier transform or a fast Fourier transform on the first acquired signal and / or the second acquired signal.

10. The method according to any one of the preceding claims, wherein, The steps of obtaining the concentration of the substance or a change in the concentration of the substance include at least one of the following: - Obtain the extracted portion of the noise spectrum and the standard deviation of the reference value; - Obtain the signal-to-noise ratio or the reciprocal of the signal-to-noise ratio of the extracted portion of the noise spectrum and the reference value.

11. The method according to any one of the preceding claims, wherein, The step of obtaining the concentration of the substance or a change in the concentration of the substance includes: quantifying the deviation from a shot noise limiting system, the shot noise limiting system being defined by the ratio of the standard deviation of an estimated portion of the noise spectrum.

12. The method according to any one of the preceding claims further comprises: - Calculate the modulation-to-noise ratio, which is given by the AC portion relative to the estimated noise, thereby quantifying the over-modulation of the signal due to the heartbeat relative to the system noise.

13. The method according to any one of the preceding claims, wherein, The liquid is blood, and the modulation of the volume and / or pressure of the blood is defined by the heartbeat.

14. The method of claim 13, further comprising: - Calculate heart rate based on the AC portion identified in the spectrum; and / or - Obtain blood glucose correlation from changes in heart rate.

15. A detector device for determining the concentration of a substance in a sample comprising a liquid containing particles, particularly glucose in blood, wherein, The refractive index of the liquid depends on the concentration of the substance dissolved in the liquid, and the particle density in the liquid is substantially constant. The detector device includes: - At least one light source and at least one detection component, wherein the at least one detection component is optically separate from the at least one light source; - Wherein, the at least one light source is configured to emit light through an exit window onto the sample; and - The detection component is configured to detect the light component corresponding to the emitted light that has been scattered through it; - A control circuit, coupled to the at least one light source and the at least one detection component; The detector device is configured to perform the method according to any one of the preceding claims.

16. The detector apparatus according to claim 15, wherein, The control circuit is configured to control the at least one light source to emit light signals during a first time period at multiple different consecutive times and to obtain signals from the at least one detection component.

17. The detector device according to claim 15 or 16, wherein, The detector device includes a plurality of photodetectors arranged in a ring or quadrilateral shape and optionally, particularly, at different distances from the at least one light source, especially centrally arranged.

18. The detector device according to any one of claims 15 to 17, wherein, The at least one light source includes a plurality of optoelectronic devices arranged at different distances from the at least one detection element; and optionally includes an annular or quadrilateral shape, optionally arranged around the at least one detection element, particularly centrally.

19. The detector device according to any one of claims 15 to 18, wherein, The at least one light source is configured to emit light of different wavelengths.

20. The detector device according to any one of claims 15 to 19, wherein, The at least one detection component includes a filter, the filter having low transmittance in a spectrum different from the spectrum emitted by the at least one light source.

Images