Micro wearable wristwatch-type non-invasive blood glucose monitoring system based on Raman spectroscopy

Abstract

The present invention is a micro wearable wristwatch-type non-invasive blood glucose monitoring system based on Raman spectroscopy, which relates to a wearable wristwatch-type system for non-invasive and rapid detection of human blood glucose concentration, and belongs to the field of non-invasive detection of blood glucose. The system integrates a near-infrared ultra-narrow bandwidth light source with a pigtail output, an avalanche photodiode (APD) for photoelectric conversion, and data processing and display into a wearable wristwatch, and is equipped with a test chamber of suitable size and ergonomics for finger fixation and Raman signal acquisition. Statistical algorithms are used to mine blood glucose Raman characteristic peaks and physical quantities related to blood glucose concentration, establish a linear relationship between spectral line peak intensity and concentration, optimize the data acquisition system and deconvolution algorithm, obtain high-precision blood glucose prediction results, and provide possibilities for clinical applications. The present invention will greatly simplify the steps of existing blood glucose detection, and has the advantages of high detection accuracy, simple operation, and wearability, and is of great significance in the field of wearable non-invasive blood glucose detection.

Description

Technical Field

[0001] The invention relates to a miniature wearable wristwatch-type non-invasive blood sugar monitoring system based on Raman spectroscopy, belonging to the field of non-invasive blood sugar detection. Background Art

[0002] In recent years, with the rise of diabetes worldwide, more and more subjects are suffering from pain and infection caused by the invasiveness of mainstream commercial blood glucose meters. Diabetes is one of the most common lifelong chronic diseases in humans. It is mainly caused by human genetic factors, immune disorders and other factors, leading to decreased pancreatic function and insulin resistance, etc., resulting in an imbalance of glucose levels in the body, manifested as abnormal glucose metabolism and hyperglycemia. Diabetes is divided into type 1 and type 2: insufficient insulin secretion in the pancreas can lead to type 1 diabetes; on the other hand, type 2 diabetes is mainly due to insulin resistance and insulin sensitivity caused by inefficient insulin use. Diabetes has a high incidence, many complications, a wide range of causes, and is difficult to cure, causing serious harm to human health.

[0003] For a long time, improving the diagnosis and treatment experience of diabetic patients has been a common goal pursued by all sectors of the industry. From the perspective of blood sugar testing, the current detection methods have gone through three technical development stages: point-of-care testing (POCT), continuous monitoring (CGM) and non-invasive monitoring. Therefore, non-invasive blood sugar monitoring has become a general trend, and non-invasive blood sugar dynamic detection technology has always been one of the research hotspots in the field of medical engineering, and it is also a challenging global problem. Since 1982, the first wrist-mounted non-invasive blood sugar meter has appeared. Today, more than 40 years later, only one non-invasive blood sugar meter has passed FDA certification, but it has long been discontinued. Only a few have passed CE certification, and the accuracy of their products has not strictly met the standards. The error rate of 15% has become the limit that non-invasive blood sugar cannot cross. Due to the low detection accuracy and poor stability, non-invasive blood sugar monitoring technology cannot be further applied. Therefore, it is urgent to develop new spectral technologies to solve the above problems.

[0004] Blood glucose detection based on optical methods is currently the mainstream non-invasive detection method. Among them, the characteristic spectrum of glucose under near-infrared (0.6μm-2.5μm) and mid-infrared (2.5μm-16μm) irradiation is the most obvious. The advantage of near-infrared is that its tissue penetration is stronger than mid-infrared, and the disadvantage is that the spectral specificity of glucose to near-infrared is not as strong as that of mid-infrared. Mid-infrared is known as the "fingerprint" in the spectrum and has strong specificity. However, due to the high requirements of mid-infrared on the emission device and the fact that it can basically not penetrate human tissue, mid-infrared has not made much progress in the field of non-invasive blood glucose monitoring. Near-infrared is currently one of the most explored technical routes in the field of non-invasive blood glucose monitoring. Therefore, we use the 784nm / 785nm dual-wavelength ultra-narrow linewidth light source in the skin window band of the near-infrared to measure the blood glucose content in the human body by measuring the Raman spectrum reflected / scattered back by the tissue, rather than the absorption method that penetrates the tissue. By irradiating the tissue with near-infrared light, the characteristic spectral information related to the vibration of blood glucose molecules is obtained, and the blood glucose concentration is obtained by analysis. Raman spectroscopy technology, as a rapid non-destructive detection technology that carries molecular "fingerprint" information, has become the preferred technology of the present invention. Although Raman spectroscopy can provide specific peak positions of vibration of specific blood glucose molecules and intensity information related to concentration, it also has the following common technical difficulties:

[0005] 1) Raman scattering light intensity is only 10 of Rayleigh scattering light -6 -10 -8 The level is therefore very susceptible to the influence of Rayleigh scattered light and spatial stray light;

[0006] 2) Limited by the working accuracy and stability of existing Raman instruments, the measurement repeatability and accuracy are far from meeting the industry standards for non-invasive blood glucose monitoring;

[0007] 3) The existing Raman spectroscopy testing system is bulky and expensive, which is not suitable for wearable integrated applications, and cannot be mass-produced;

[0008] 4) Existing Raman testing systems mostly work independently or in conjunction with other spectral technologies and have not yet been applied in medical devices, especially home medical devices.

[0009] In view of the above common technical difficulties, the present invention proposes a high-precision Raman spectroscopy non-invasive blood glucose level acquisition method, the innovation of which lies in: selecting a vertical cavity surface laser emitter (Vcsel) as a miniature dual-wavelength light source, selecting an ultra-narrow bandwidth filter to replace the traditional grating spectrometer, selecting an avalanche photodiode to replace the traditional charge-coupled device (CCD), and integrating them into a wearable watch, and equipping it with an ergonomic finger test chamber to assist in Raman signal acquisition to complete the detection of human blood glucose level. Summary of the invention

[0010] The purpose of the present invention is to realize non-invasive and accurate detection of blood sugar levels based on wearable devices, and to propose a micro wearable wristwatch-type non-invasive blood sugar monitoring system based on Raman spectroscopy, so that users can perform self-blood sugar monitoring in a portable, frequent and rapid manner.

[0011] The purpose of the present invention is achieved through the following technical solutions.

[0012] The dual-wavelength ultra-narrow bandwidth near-infrared light source laser is transmitted to the test chamber through the adapter via the stretchable bare fiber placed on the rotatable optical fiber disk. The test chamber is equipped with a lens group that can move in the vertical dimension. By controlling the vertical dimension distance, the focus is accurately focused on the lower surface of the fingernail fixed according to the nail gap. The light source is triggered by a specific timing to stimulate the abundant capillaries under the nail to obtain the human blood glucose Raman signal, and two groups of human Raman scattered light with a slight frequency shift are obtained. The dichroic mirror in the test chamber reflects the signal generated by the excitation to the reflector, and further transmits it to the avalanche photodiode in the watch through the optical fiber to complete the photoelectric signal conversion. Finally, the relationship between the Raman spectrum and the blood glucose concentration is established through an intelligent algorithm to accurately predict the human blood glucose level and display it on the watch screen.

[0013] The dual-wavelength light source is a vertical cavity surface emitting laser (Vcsel), which can independently emit two laser beams with similar wavelengths, large output power and narrow laser spectrum line width, preferably but not limited to 784nm and 785nm. In particular, the excitation light wavelength is selected by using the pigtail output of the built-in Bragg grating to obtain excitation light with a central wavelength of 784nm and 785nm with a line width of ≤0.01nm, and its output power is ≥1w, and the two wavelength lasers can be triggered according to a certain timing to act on human nails respectively.

[0014] As an alternative, 1064 nm excitation light that effectively avoids fluorescence effect can be selected as an ultra-narrow linewidth single-wavelength light source, which can effectively avoid the fluorescence background that annihilates the Raman signal.

[0015] The optical fiber adopts bare fiber without armor, which can effectively reduce the space and can be stretched according to the required length. It is formed by the convergence of three pigtails of dual-wavelength light source and avalanche photodiode, and has a certain toughness to prevent easy breakage.

[0016] The optical fiber reel is an automatically retractable reel, the center of which coincides with the center of the watch body base, and can rotate around the central axis according to the tension. The advantage of using this device is that the optical fiber can be automatically retracted into the watch when not testing, and can be kept at a specific length during testing, which plays a role in protecting the optical fiber and stabilizing the test.

[0017] The adapter contains an excitation end and a collection end optical fiber coupling end, wherein the optical fiber connected to the excitation light source is coupled to the excitation end, and the optical fiber connected to the avalanche photodiode is coupled to the collection end; when not testing, the adapter is fixed to one side of the watch through a card slot. During testing, the adapter can be unplugged and the optical fiber can be stably fixed to the test chamber through the card slot above the test chamber, thereby completing the combination of the watch and the test chamber.

[0018] The test chamber mainly includes an excitation light fiber coupler, an excitation light beam expansion collimating lens group, a dichroic mirror, a focusing lens group, a finger slot, a finger hole, a high reflective mirror and a collecting fiber coupler. In particular, the test chamber should be in an absolutely dark environment during the test to ensure that the test will not be affected by ambient stray light.

[0019] The excitation light fiber coupler is connected to the excitation end fiber coupling port in the transfer port to ensure that the excitation light enters the test chamber while achieving maximum coupling efficiency.

[0020] The excitation light expansion and collimation lens group is used to expand the light output through the excitation light fiber coupler and collimate it into spatial light, and then it is incident on the focusing lens group.

[0021] The focusing lens group can move in the vertical dimension and automatically adjust its position by collecting information from signal feedback, thereby ensuring that the collimated spatial light can be accurately focused on the lower surface of the human nail and stimulate the abundant capillaries in the nail bed.

[0022] The finger slot is used to place and fix the finger. In order to be suitable for different groups of people, the inner sheath is designed to fit tightly around the finger, ensuring the comfort of the finger while preventing stray light from entering the test chamber. In particular, a nail slit is designed in the finger slot, and the protruding part of the nail can be inserted into the nail slit to fix it, overcoming the problem of inaccurate testing caused by focus changes due to slight movements of the finger.

[0023] The finger hole serves to place a finger into the test chamber, and is protected by a small cover when not in use to prevent dust from falling into the test chamber.

[0024] The high-reflection mirror is used to reflect other irrelevant light such as Raman scattered light and Rayleigh scattered light excited from the human nail reflected by the dichroic mirror.

[0025] The collecting fiber coupler is used to focus and couple the Raman scattered light and Rayleigh scattered light and other irrelevant light excited from the human nails, and is connected to the collecting fiber coupling end of the adapter, so that the collected light enters the avalanche photodiode through the optical fiber.

[0026] The avalanche photodiode is preferably but not limited to a silicon-based avalanche photodetector, which has an amplifier and a light detector in the same package. During system installation, it is required to be in an absolutely dark environment to reduce environmental noise and parasitic capacitance. Its detection sensitivity can reach 0.52A / W, and its maximum dark current is 2nA. It is used to detect the two groups of specific small-band Raman scattered light signals with slight frequency shifts and convert them into electrical signals, thereby obtaining two groups of spectra.

[0027] In particular, a narrow bandwidth filter is installed at the front end of the APD to filter out the Rayleigh scattered light, spatial stray light and irrelevant Raman scattered light other than the Raman characteristic peak of the specific band of human blood glucose; according to the formula (10 7 / (10 7 / λ-v)) to calculate the central wavelength of the filter, wherein λ is the wavelength of the micro light source, and v is the Raman characteristic peak of human blood glucose. In the present invention, v is preferably 1125cm, which has a large correlation with blood glucose content and is not easily affected by objective factors such as human heat metabolism. -1 The single Raman characteristic peak at 784nm and 785nm can be obtained by excitation, and the wavelengths corresponding to the characteristic peaks of blood glucose are 859.8nm and 861nm respectively, that is, the transmittance of the filter to the wavelength band of 859.8nm to 861nm is ≥90%, and the transmittance to the light of the other wavelength bands is <0.0001%;

[0028] Beneficial Effects of the Invention

[0029] It can be seen from the above technical solutions that the micro wearable wristwatch-type non-invasive blood glucose monitoring system based on Raman spectroscopy disclosed in the present invention has at least the following beneficial effects:

[0030] 1) A unique and effective nail surface signal acquisition scheme has been set up. Although the nail itself does not contain capillaries, the nail bed it covers contains a large number of capillaries. The nail is composed of 3-4 layers of hard scaly stratum corneum overlapping, and the light transmittance reaches about 70% (actual measurement), so while ensuring that most of the laser can reach the nail bed, it also avoids laser damage to the skin;

[0031] 2) Based on the frequency-shifted excitation Raman differential spectrum, the multiple iterative deconvolution differential Raman restoration algorithm is organically combined with the partial least squares method. Relying on a huge database, the most relevant Raman peak is captured as an effective spectral signal, and the correlation between blood glucose concentration and Raman enhanced signal intensity is established. Wavelet transform and neural network algorithms and related correction models are further used to improve detection accuracy and measurement repeatability;

[0032] 3) Use a micro vertical cavity surface emitting laser as a dual-wavelength light source; propose to use an ultra-narrow bandpass filter to replace the traditional expensive grating to filter out the Rayleigh scattered light and irrelevant Raman scattered light other than the Raman characteristic peak of the human body in a specific band; propose to use an avalanche photodiode to replace the traditional CCD for Raman signal detection to complete the detection of extremely weak light signals and the conversion of photoelectric signals; and integrate them all into a wearable watch, and equip it with an ergonomic finger test chamber to assist users in quickly and accurately monitoring their blood sugar levels;

[0033] 4) The whole set of equipment is easy to operate and carry. It does not require much preparation before testing, and only requires simple cleaning of the nail surface. It has a long service life and will not cause waste of related consumables, greatly reducing costs. The whole process will not cause any discomfort or pain to the subjects, and is especially suitable for children and the elderly. It is of great significance to enhance the enthusiasm of diabetic patients for self-blood sugar monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a schematic diagram of a wristwatch of a micro wearable wristwatch-type non-invasive blood glucose monitoring system based on Raman spectroscopy of the present invention;

[0035] Figure 2 It is a schematic diagram of a finger test chamber of a micro wearable wristwatch-type non-invasive blood glucose monitoring system based on Raman spectroscopy of the present invention;

[0036] Figure 3 This is a schematic diagram of a micro wearable wristwatch-type non-invasive blood glucose monitoring system based on Raman spectroscopy in the present invention, in which a wristwatch and a finger test chamber are shared;

[0037] Among them, 1-watch body, 2-watch strap, 3-vertical cavity surface emitting 784nm laser, 4-vertical cavity surface emitting 785nm laser, 5-bare fiber, 6-automatic retractable winding fiber disk, 7-avalanche photodiode, 8-adapter, 9-fiber built-in Bragg grating, 10-ultra-narrow bandwidth filter, 11-finger test chamber body, 12-excitation light fiber coupler, 13-adapter card slot, 14-excitation end beam expansion and collimation lens group, 15-dichroic mirror, 16-vertically movable focusing lens group, 17-nail seam, 18-finger groove, 19-finger hole, 20-high reflector, 21-collection end focusing lens group, 22-collection light fiber coupler, 23-stretchable fiber. DETAILED DESCRIPTION

[0038] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0039] The present invention proposes a miniature wearable wristwatch-type non-invasive blood glucose monitoring system based on Raman spectroscopy. A dual-wavelength ultra-narrow bandwidth near-infrared light source laser is transmitted to a test chamber through a stretchable bare fiber placed on a rotatable optical fiber disk through an adapter. A lens group that can move in the vertical dimension is provided in the test chamber. The focus is accurately focused on the lower surface of the fingernail fixed according to the nail gap by controlling the vertical dimension distance. The light source is triggered by a specific timing to stimulate the abundant capillaries under the nail to obtain the human blood glucose Raman signal, and two groups of human Raman scattered light with a slight frequency shift are obtained. The dichroic mirror in the test chamber reflects the signal generated by the excitation to the reflector, which is further transmitted to the avalanche photodiode in the watch through the optical fiber to complete the photoelectric signal conversion. Finally, the relationship between the Raman spectrum and the blood glucose concentration is established through an intelligent algorithm to achieve accurate prediction of the human blood glucose level and display it on the watch screen.

[0040] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments and with reference to the accompanying drawings. However, the present invention can be implemented in different forms and should not be construed as being limited to the embodiments set forth herein.

[0041] like Figure 3 As shown, this embodiment discloses a wearable non-invasive ancestry detection device that uses a wristwatch in combination with a finger test chamber.

[0042] As an optional embodiment, when not testing, the test chamber is separated from the watch, and the watch can work normally. During the test, first remove the side adapter 8 of the watch, and pull a specific length to fix it in the test chamber card slot, then select the blood sugar detection function through the watch button, at this time, the 784nm laser 3 and the 785nm laser 4 are triggered, and they work independently for 20s, which are used to excite the Rayleigh scattered light and Raman scattered light generated by the capillaries on the nail bed under the surface of the human nail. In particular, the above-mentioned laser has a pigtail output, and the output light width is compressed to ≤0.01nm through the built-in Bragg grating 9. The excitation light reaches the excitation light coupling end of the adapter 8 through the bare fiber 5 placed on the automatic retractable winding fiber optic disc 6, and enters the test chamber dichroic mirror 15 after connecting with the finger test chamber 11. The dichroic mirror is highly transparent to the narrow line width 784nm and 785nm excitation light, and transmits the excitation light to the focusing lens group 16 that can move in the vertical dimension, and the laser focus position can be accurately controlled by controlling the moving distance. The test finger enters the test chamber through the finger hole 19. The inner surface of the finger groove 18 is composed of a sheath, which can tighten and fix the finger and ensure the dark environment of the test chamber. In particular, the front end of the finger groove is used to further fix the finger by the nail slit 17. The protruding part of the test finger nail can be inserted into the nail slit to prevent the finger from moving up and down during the test, causing test errors caused by focus changes. The Raman scattered light, Rayleigh scattered light and other irrelevant stray light generated by the excitation from a large number of capillaries on the nail bed under the human nail are collected by the focusing lens group 16, and reflected to the high reflective mirror 20 through the dichroic mirror 15, and finally reach the collection end focusing lens group 21. Through the connection relationship, it is finally coupled to the collection optical fiber at the optical fiber coupler at the collection end of the transfer interface, and finally reaches the avalanche photodiode 7. In particular, there is a built-in narrow bandwidth filter 10 at the front end of the avalanche photodiode to filter out the above-mentioned scattered light and filter out the specific small-band Raman characteristic peak of the human body (1125cm -1 ) The corresponding wavelength transmittance is ≥90%, and the transmittance of scattered light and spatial stray light in other bands is <0.0001%. There are only two groups of small bands with slight frequency shift, including 1125cm -1 The light from the Raman characteristic peak of human blood sugar enters the APD, completes the detection of extremely weak Raman scattering signals and the conversion of photoelectric signals, and finally uses a dedicated processor for A / D conversion. The optional DSP is used to receive, process and reconstruct digital signals, realize calibration correction, data visualization and other processing, and display the final predicted value on the watch screen to complete the entire test process.

[0043] The specific description above further illustrates the purpose, technical solutions and beneficial effects of the invention in detail. It should be understood that the above is only a specific implementation of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A micro wearable wristwatch-type non-invasive blood glucose monitoring system based on Raman spectroscopy, characterized by: The dual-wavelength ultra-narrow bandwidth near-infrared light source is transmitted to the test chamber through the stretchable bare fiber placed on the rotatable optical fiber disk through the adapter. The test chamber is equipped with a lens group that can move in the vertical dimension. By controlling the vertical dimension distance, the focus is accurately focused on the lower surface of the fingernail fixed according to the nail seam. The light source is triggered by a specific timing to stimulate the abundant capillaries under the nail to obtain the human blood glucose Raman signal, and two groups of human Raman scattered light with a slight frequency shift are obtained. The dichroic mirror in the test chamber reflects the excited signal to the reflector, and further transmits it to the avalanche photodiode in the watch through the optical fiber to complete the photoelectric signal conversion. Finally, the relationship between the Raman spectrum and the blood glucose concentration is established through the intelligent algorithm to achieve accurate prediction of the human blood glucose level and display it on the watch screen. The dual-wavelength ultra-narrow bandwidth near-infrared light source adopts a vertical cavity surface emitting laser, and selects 784nm and 785nm ultra-narrow bandwidth lasers with a bandwidth of ≤0.01nm through a Bragg grating with a length of 10mm built into the pigtail, with an output power of ≥1.25w, and the two wavelengths can be triggered according to a specific timing; The optical fiber is a bare fiber, which is formed by the convergence of three pigtails of a dual-wavelength light source pigtail and an avalanche photodiode. The bare fiber is placed around the optical fiber disk, and the other end is fixed to the transfer interface. The appropriate length can be selected according to the tension rotation; the optical fiber disk is an automatic retractable winding disk, and the center of the base coincides with the center of the watch; The adapter is divided into an excitation end and a collection end. The dual-wavelength light source fiber among the three optical fibers after convergence is connected to the excitation end at the adapter, and the avalanche photodiode fiber is connected to the collection end. The adapter is fixed to one side of the watch when not testing. During testing, it can be lightly pulled to a specific length and directly fixed to the top of the test chamber. The test chamber is an absolutely dark environment and includes the following parts: Interface adapter slot: used to fix the interface adapter to ensure the stability of the interface during the test; Excitation and collection fiber coupling device: used to couple the excitation end and collection end optical fibers in the adapter to the test chamber to ensure that the coupling efficiency is greater than 85%; Dichroic mirror: used for transmitting the ultra-narrow bandwidth dual-wavelength laser from the excitation end of the excitation and collection optical fiber coupling device, and reflecting glucose Raman scattering and other irrelevant signals from the capillaries covered by human nails; Lens group: used to focus the excitation light to the capillaries under the nail surface, and collect the generated Rayleigh scattered light and Raman scattered light; Finger slot: used to put fingers in and ensure a dark environment in the test chamber; Nail seam: used to insert the protruding part of the nail to fix the human finger so that it will not move or change position at will, thus ensuring the stability of the data collection during the test; Reflector: used to reflect the glucose Raman scattered light and other irrelevant signals from the capillaries covered by the human nails reflected by the dichroic mirror to the collecting end of the excitation and collection optical fiber coupling device.

2. The micro wearable wristwatch type non-invasive blood glucose monitoring system based on Raman spectroscopy as claimed in claim 1, characterized in that: The avalanche photodiode has a pigtail input, and a narrow bandpass filter is attached to the front end of the photodiode. The filter is used to filter out Raman scattered light and Rayleigh scattered light outside the specific band of human body Raman scattered light. It is characterized in that: the transmittance of Raman scattered light in the specific band is greater than or equal to 90%; the transmittance of scattered light and Rayleigh scattered light outside the specific band is less than 0.0001%.

3. According to claim 1 or 2, the micro wearable wristwatch type non-invasive blood glucose monitoring system based on Raman spectroscopy, wherein the two groups of human Raman scattered light with a slight frequency shift are determined according to the inverse of the difference between the frequency of the two incident light beams and the scattered light frequency; the scattered light is characterized by: the human blood glucose Raman characteristic peak 1125cm -1 The influence of metabolic heat and human body functions on intensity can be ignored, and the change in intensity can be considered to be caused by a single variable of blood glucose concentration.

4. The micro wearable wristwatch type non-invasive blood glucose monitoring system based on Raman spectroscopy as claimed in claim 1, characterized in that: Data processing and display uses the built-in multiple iterative deconvolution differential Raman algorithm to remove the background, extract, amplify and restore the signals of the two sets of spectra, and predict the obtained Raman characteristic peak intensity using the built-in partial least squares regression model, and display the predicted values ​​on the screen.

5. The micro wearable wristwatch type non-invasive blood glucose monitoring system based on Raman spectroscopy according to claim 1, characterized in that: Integrating non-invasive blood glucose testing into a watch can enable frequent, non-invasive, accurate and portable blood glucose testing.

Images