WO2023280017A1

WO2023280017A1 - Non-invasive blood glucose detector and detection method

Info

Publication number: WO2023280017A1
Application number: PCT/CN2022/101907
Authority: WO
Inventors: 刘炜
Original assignee: 无锡轲虎医疗科技有限责任公司
Priority date: 2020-09-11
Filing date: 2022-06-28
Publication date: 2023-01-12
Also published as: CN113261954B; CN113261954A; CN112022167A

Abstract

A non-invasive blood glucose meter comprises a housing (1), a controller (2), a power supply module (3), cooling fins (4), a power switch (5), an LED light source (6), and a spectral sensor (7). A finger compartment (8) is provided in the housing (1); the power supply module (3) and the controller (2) are electrically connected to each other and are disposed in the housing (1); the LED light source (6) is disposed at the end of the finger compartment (8) and is located on the top surface of the finger compartment (8); the cooling fins (4) are connected to the spectral sensor (7); the spectral sensor (7) is located at the end of the finger compartment (8) and located on the bottom surface of the finger compartment (8); the LED light source (6) and the spectral sensor (7) are located on a same straight line. During non-invasive blood glucose detection, position limit, classification and recognition, detection position determining, and fingertip pressing degree determining can be performed on the finger before non-invasive blood glucose detection, and most stable spectral detection is performed on the finger when all interference factors are minimized.

Description

Non-invasive blood glucose detector and detection method

This application is based on and claims the priority of Chinese Patent Application No. 202110778786.X filed at the State Intellectual Property Office of China on July 9, 2021, the entire contents of which are hereby incorporated by reference.

technical field

The invention relates to the improvement of non-invasive blood sugar detection technology, in particular to a non-invasive blood sugar detection instrument and detection method with multi-sensing and highly accurate data.

Background technique

Blood glucose testing is a key link in diabetes treatment, but traditional testing requires blood sampling, and the trauma of blood sampling makes it difficult for patients to adhere to daily monitoring of blood glucose. This is also a problem that has plagued the medical profession for many years.

Near-infrared spectrum detection of blood sugar technology has been expected to achieve non-invasive detection, but the main reasons that hinder the use of near-infrared spectrum detection of blood sugar technology to achieve non-invasive detection are as follows.

1. Weak signal

In human blood, more than 90% of the composition is water, and the proportion of blood is only 7-8%, and the content of blood sugar in blood is very low. Moreover, the absorption of near-infrared light by water is relatively serious, which will seriously interfere with non-invasive detection. In addition, the core of the near-infrared spectral band is the double frequency and combined frequency absorption of molecules. The absorption peaks are broad and overlap each other seriously. The magnitude of absorbance is orders of magnitude different from the fundamental frequency of mid-infrared. To sum up the above factors, if we want to accurately detect the change information of the weak blood sugar components, we will put forward higher requirements on the performance of the spectrum acquisition system.

2. Background interference

Human tissues such as skin, muscle, and bone are strong near-infrared absorbers, and the human body spectrum will carry a large amount of interference information related to tissues, and the effective information that can be used for analysis is easily submerged in the strong background. Therefore, the problem of tissue background interference is one of the important reasons affecting the accuracy of non-invasive blood glucose detection.

3. Individual differences

There are large differences in tissue characteristics such as blood, skin, and muscle among different individuals, and even different tissue background components in different parts of the same individual, which will complicate the background noise of the collected spectrum, and further increase the extraction of blood component information from human spectra. difficulty.

4. Changes in blood flow volume

The human body is a complex living body. Physiological phenomena such as heartbeat and blood circulation will cause periodic fluctuations in blood flow volume. The time-varying characteristics of blood flow volume will cause changes in the absorbance of the human body in the near-infrared spectrum, and have a significant impact on the measurement results. Obvious impact, mainly manifested as instability in the spectral time domain.

5. The wavelength range of the photoelectric sensor is too wide

The wavelength receiving range of the photoelectric sensor that can be used for blood sugar detection wavelength is too wide to accurately receive the spectral information of a specific wavelength LED. It is necessary to use a filter to cut off the wavelength receiving range to make it narrower. After using the filter, the light transmission rate will be reduced. Insufficient, affect the accuracy.

6. LED power is too small, insufficient light transmission

The power of specific wavelength LEDs on the market that can be used for blood sugar detection is too small, concentrated at 1-3mw, which makes the pass rate of the light illuminated to the finger insufficient.

Just as the above-mentioned six reasons are weak signal, background interference, individual differences, changes in blood flow volume, too wide wavelength range of photoelectric sensor, too small LED power, and insufficient light transmission, so how to extract effective information from strong background spectrum is a near-term challenge. The key issues that need to be solved in the non-invasive detection of blood glucose by infrared spectroscopy.

At present, Chinese patent publication CN108593593A discloses a serial dual-infrared spectrum non-invasive blood glucose measurement device. The whole device is composed of a broadband infrared light source, a measurement hole, a double filter switcher, an infrared photoelectric sensor, and a signal acquisition and processing circuit. The infrared photoelectric sensor is located behind the double-filter switcher and can convert spectral energy information into corresponding voltage signals.

The above scheme uses a broadband infrared light source and distributes the spectral energy in the near-infrared (800nm-1100nm) or short-wave infrared (1000-1800nm) spectral range. The infrared photoelectric sensor has near-infrared (800nm-1100nm) or short-wave ) spectral sensitivity. The infrared photoelectric sensor is located behind the double-filter switcher, and analyzes the wavelength carrying the blood sugar level information in the form of receiving the effective wavelength after filtering out the invalid wavelength. Although the effective wavelength carries blood glucose level information, it is still uncertain where the specific wavelengths of the effective blood glucose level information are concentrated. Usually, it is necessary to use an algorithm for conversion, combined with big data to determine the specific blood glucose level.

At present, this solution is the mainstream solution for blood sugar level detection. The main problem is that the wavelength range used is wide, and there are many invalid information mixed in the wavelength. It relies too much on algorithms and big data, and algorithms and big data require effective information. When there is a lot of invalid information, the accuracy of the results obtained by using algorithms and big data is still low.

Some technical R&D personnel have also discovered this problem, so when using spectral measurement to detect blood sugar, designers also focus on narrowing the wavelength of the light source, so that the detected wavelength is an effective wavelength, and the wavelength range is relatively narrow , so as to filter out invalid optical information as much as possible.

Chinese patent application publication No. CN112022167A discloses a non-invasive blood sugar detection method based on a spectral sensor, which includes the following steps: 1) a spectral sensor is designed at the fingertip position, and an LED is designed on the other side of the fingertip position; 2) Adapt the Fabry-Perot interferometer tunable filter in the spectral sensor, and adjust the optical receiving range of the tunable filter to nm level; Spectral sensor for collection; 4): The light emitted by 1720nm LED passes through human tissue and then collected by 1550nm-1850nm spectral sensor.

The blood glucose level measured during the actual use of this solution is basically the same as the detection result of the invasive blood glucose level under normal circumstances. However, a problem was also found in the actual use process, that is, when the fingertips of different fingers are sent to the non-invasive blood glucose meter for detection, the measured value has a certain deviation. When the same finger is sent to the non-invasive blood glucose meter for testing, due to the different strength of the fingertips when it is inserted into the bottom and the existence of fingernails, the measured value also has a certain deviation.

To sum up, it can be seen that how to further reduce the error of spectroscopic technology to realize non-invasive blood glucose detection is a technical problem to be solved.

Contents of the invention

The purpose of this application is to provide a multi-sensory and highly accurate non-invasive blood glucose detector and detection method, which can limit, classify and identify, determine the detection position and squeeze the fingertips before performing non-invasive blood glucose detection on fingers Judging by the degree, the most stable spectrum detection is performed on the finger when all the interference factors are reduced to the minimum.

This application provides the following technical solution: a non-invasive blood glucose detector with multi-sensory and highly accurate data, including a housing, a controller, a power module, a heat sink, a power switch, an LED light source, and a spectral sensor. A finger compartment is provided, the power supply module is electrically connected to the controller and both are arranged in the casing, the LED light source is arranged at the end of the finger compartment and is located on the top surface of the finger compartment, the heat sink is connected to the spectrum sensor, and the The spectrum sensor is located at the end of the finger compartment and on the bottom surface of the finger compartment, the LED light source and the spectrum sensor are located on the same straight line, the end of the finger compartment is provided with a notch, and the edge of the notch is provided with a slope stop block, the slope block is provided with a plurality of temperature sensors side by side, the temperature sensor is electrically connected to the controller, the slope block is provided with a through hole for detecting the passage of light, and also includes fingertip pressure Detection mechanism, the fingertip pressure detection mechanism includes a detection bracket, a fingertip stopper and a pressure sensor, the detection bracket is arranged on the housing, and the fingertip stopper is connected to the detection bracket through a pressure sensor; the pressure sensor is located at At the notch and at the edge of the slope block, the pressure sensor is also electrically connected to the controller.

In some embodiments, the detection bracket is detachably connected to the housing, and the detection bracket also abuts against the controller.

In some embodiments, a guide spring is further included, the guide spring is nested on the pressure sensor, and the guide spring and the pressure sensor are inclined.

In some embodiments, a guide rail groove is provided on the side wall of the finger compartment, and a guide rail is provided on the fingertip stopper, and the fingertip stopper slides and fits with the guide rail groove through the guide rail.

In some embodiments, the fingertip stopper is provided with a fingertip limiting groove with a width of 5 mm to 1 cm.

In some embodiments, the top of the fingertip stop is arc-shaped, and the height of the notch is greater than the height of the fingertip stop.

In some embodiments, it also includes a semiconductor cooling sheet and a heat dissipation copper sheet. The spectral sensor is also connected to the heat dissipation copper sheet. To the heat dissipation copper sheet, the heating end of the semiconductor refrigeration sheet is attached to the heat dissipation sheet.

On the other hand, a multi-sensory and highly accurate non-invasive blood glucose detection method comprises the following steps:

1) Turn on the power switch, and the power module powers on the LED light source, spectrum sensor, temperature sensor and pressure sensor respectively through the controller;

2) Insert your finger into the finger compartment and probe into the bottom of the finger compartment. The nail part of the fingertip will be located in the gap between the fingertip block and the notch. The front end of the fingertip is limited by the fingertip limit groove and makes the finger The tip is horizontal, and the guide spring provides damping effect and limits the forward displacement of the fingertip;

3) Divide the fingers into three categories, the first category is the thumb, the second category is the index finger, middle finger and ring finger, and the third category is the little finger;

4) When the pressure sensor receives the pressure signal and remains stable, it is determined that the fingertip has reached the bottom of the finger compartment and remains horizontal, and the pressure value received by the pressure sensor is a coefficient K; when the temperature received by the temperature sensor changes, it is determined that the fingertip Temperature T1, the unchanged temperature is the ambient temperature T2, the number N of changes in the temperature sensor;

5) After the light of the specified wavelength is emitted by the LED light source, the light passes through the through hole and passes through the detection part at the front of the fingertip. The spectral sensor is equipped with a Fabry-Perot interferometer tunable filter, and the tunable filter can be adjusted Spectrum reception is performed after the optical receiving range of the device reaches the nm level;

6) According to the number N of changes in the temperature sensor, it is judged that the type of the finger belongs to the first, second or third category, combined with the coefficient K, fingertip temperature T1 and ambient temperature T2, combined with the spectral sensor to receive the light signal and convert it The ADC values are sent to the controller together, and the controller calls the algorithm in the built-in data storage module to calculate the blood glucose value.

In some embodiments, the LED light source emits specified wavelengths of 1500 nm, 1525 nm, 1550 nm and 1575 nm.

In some embodiments, when the power switch is turned on, the power module also energizes the semiconductor cooling chip through the controller, and the semiconductor cooling chip directly cools down the spectral sensor through the heat dissipation copper sheet, so that the spectral sensor can emit light at a specified temperature. induction reception.

Regarding the existing disclosed non-invasive blood glucose meter, the inventor found the main reason after repeated tests is that the spectral detection method of the non-invasive blood glucose meter is mainly aimed at the 2-5 mm of the fingertip, which has not been blocked by bones. Therefore, when the spectrum sensor is used for collection, it is possible to collect as much wavelength as possible containing effective blood sugar information. But precisely because this part belongs to the front end of the fingertip, there is also the influence of the length of the nail. Therefore, when the users of the non-invasive blood glucose meter measure, some of the parts are squeezed, and some are in a relaxed state, but the density of the finger will change after being squeezed, and the strength of the finger inserted into the non-invasive blood glucose meter Different, the situation of being squeezed is also different. There are even some ideal detection states where the fingertip to be tested is not perpendicular to the light source when the finger is inserted.

The multi-sensing and highly accurate non-invasive blood glucose detection method of the present application can use the above-mentioned non-invasive blood glucose detection instrument.

Compared with the prior art, the design of the present application adopts the design of the slope stopper, which can not only block the end of the finger to a certain extent, but also provide a tactile reminder for the finger to reach the designated position. Multiple temperature sensors can detect both the temperature of the environment and the temperature of the finger, and can also judge the width of the fingertip according to the number of temperature sensors that have changed, thereby judging the category of the penetrating finger.

There is a notch at the end of the finger compartment, and a fingertip pressure detection mechanism is set at the notch position. The guide spring of the fingertip pressure detection mechanism supports the fingertip stopper and keeps it tilted. When the tip touches the fingertip block, it can play a damping effect, and at the same time, the pressure sensor can further detect the extrusion variable of the finger end.

The height of the notch at the end of the finger compartment is higher than the height of the fingertip block. The top of the fingertip block can be arc-shaped, and the gap formed in the middle can be passed by the nail to ensure that there will be no light leakage or light leakage caused by the deformation of the nail during detection. Tissue over-squeeze problem.

The soft material layer used for the fingertip stopper is sponge foam material inside, and the outer layer is lined with soft fabric. At the same time, there is a fingertip limit groove to fine-tune the level of the fingertips, prevent fingertips from shifting, and avoid excessive fingertips. Squeeze, if there is excessive extrusion, the variable of the pressure sensor can intervene and serve as a reference for blood sugar detection.

Cooperate with the semiconductor refrigeration sheet to cool down the heat dissipation copper sheet, so as to ensure that the temperature of the spectral sensor is constant and will not cause the detection data to drift due to temperature changes. Rapid heat dissipation is performed through the heat sink and stainless steel heat dissipation etched mesh, and the cooling fan is used to dissipate heat in the housing, which can Ensure that the temperature of the detection environment is stable even when the instrument is used for a long time.

Description of drawings

Figure 1 is a map of the distribution of blood vessels in the palm;

Fig. 2 is the schematic diagram of fingertip part;

Fig. 3 is the structural representation of the detector of the present application;

Fig. 4 is the internal structural schematic diagram 1 of the structural schematic diagram of the detector of the present application;

Fig. 5 is a partially enlarged schematic diagram of schematic diagram 1 of the present application;

Fig. 6 is the internal structural schematic diagram II of the structural schematic diagram of the detector of the present application;

FIG. 7 is a partially enlarged schematic diagram of the second schematic diagram of the present application;

Fig. 8 is the explosion schematic diagram of the detector of the present application;

FIG. 9 is a block diagram of the controller of the present application.

Reference signs: 1. Housing; 2. Controller; 3. Power module; 4. Heat sink; 5. Power switch; 6. LED light source; 7. Spectrum sensor; 8. Finger compartment; 9. Notch; 10 , Slope stopper; 11. Temperature sensor; 12. Through hole; 13. Fingertip pressure detection mechanism; 14. Detection bracket; 15. Fingertip stopper; 16. Pressure sensor; 17. Guide spring; 18. Rail groove ;19, guide rail; 20, fingertip limit groove; 21, indicator light; 22, heat dissipation hole; 23, upper cover; 24, buckle ring; 25, fixed bracket; 26, fixed plate; 27, stainless steel heat dissipation etching net; 28. Rubber sealing ring; 29. Cooling fan; 30. Fixing bracket for cooling fan; 31. Copper sheet for cooling; 32. Semiconductor cooling chip.

detailed description

The "embodiment", "an embodiment", "another embodiment" or "certain embodiments" mentioned in the specification refer to the described specifically related features and structures related to the embodiment or characteristics are included in at least one embodiment. Thus, "an embodiment", "an embodiment", "another embodiment" or "certain embodiments" are not necessarily all referring to the same embodiment. Furthermore, the particular features, structures or characteristics may be combined in any manner in one or more embodiments.

The preferred embodiments of the present application will be described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described here are only used to illustrate and explain the present application, and are not intended to limit the present application.

As shown in Figure 1, the distribution of blood vessels in the palm. Observing the figure, we can see that there are rich capillary networks distributed on the fingertips, which can effectively reflect the blood sugar content in the human body, and this position has obvious characteristics of blood flow volume changes. The muscle and bone tissues of fingers are relatively thin, so the influence of background noise information is relatively small. In addition, the tip of the finger is convenient for measurement, and there is no psychological burden on the subject, which is conducive to obtaining a stable spectral signal with a high signal-to-noise ratio. All non-invasive blood glucose meters in the prior art have a fixed size and shape of the finger compartment, and the installation positions of the spectral sensor and the LED light source in the finger compartment are fixed, although the device manual will specify which finger to use for detection Make advance instructions, usually the index finger and middle finger. As shown in Figure 2, after the light passes through the detection part of the fingertip, the light is collected by the spectral sensor. If the size, direction and extrusion degree of the finger are in the state shown in Figure 2 every time the detection is performed, the detection site is determined and kept uniform, which is the most ideal state. But in actual use, everyone's usage habits are different, and it is easy to ignore this point. At the same time, the non-invasive blood glucose meter does not necessarily remain horizontal during the use of each person, nor does it necessarily remain horizontal when the finger is inserted into the bottom of the finger compartment. It is determined that there are many variables before the measurement, and these variables are the influencing variables of the non-invasive blood glucose detection.

The inventor found the following three main problems in the use of non-invasive blood glucose meter equipment during research and development:

First, it is impossible to determine which finger is inserted into the finger compartment. The shape and width of the fingertips of the index finger, middle finger and ring finger are similar, but the width and length of the little finger and the thumb are quite different. When a finger is inserted into the finger compartment, the degree of extrusion of the fingertips of different finger types and lengths is naturally uncertain, which in turn will affect the extrusion degree of the detection part, which will affect the LED light source. Wavelength change and light intensity attenuation after penetrating the detection site.

Second, the fingertip nails of each finger vary in length according to individual habits. When a finger is inserted into the finger compartment, the nail on the fingertip will contact the front end of the finger compartment, forming deformation and causing light leakage, and causing the detection part to move backward or forward. The ideal position of the detection site is deviated from the position of the test light source, and the deformation of the nail will also cause changes in the tissue density of the actual detection site.

Third, when each finger is inserted into the finger compartment, the fingertip does not necessarily remain horizontal. During the insertion process, the fingertip will have a certain inclination, and the test site and tissue density after the inclination will also be different from the preset value. deviation.

The technology of non-invasive blood glucose meters has certain resistance in the process of popularization and advancement, partly because the devices currently on the market based on near-infrared spectroscopy for the development of non-invasive blood glucose are all designed based on specific wavelength ranges or specific wavelength LEDs plus traditional photoelectric sensors. Due to the wide range of receiving wavelengths of the photoelectric sensor, if you want to obtain high-precision optical data, you need to use a filter to cut off the wavelength. However, due to the influence of the processing precision of the optical filter, the precision of the optical signal is insufficient.

On the other hand, due to individual differences, different user habits, and solidified structural design, although it needs to be calibrated before use, but adapted to individual characteristics, the above problems will exist in each independent test. In order to solve the above problems, the inventor designed a non-invasive blood glucose detector with multi-sensory and highly accurate data, and established a new fingertip non-invasive blood glucose measurement method.

The multi-sensory and highly accurate non-invasive blood glucose detector of the present application is shown in FIG. 8 during specific assembly. The power switch 5 is installed on the upper cover 23 , and the power switch 5 is electrically connected to the controller 2 . The upper cover 23 is detachably connected to the housing 1 through a buckle 24 . A cooling hole 22 is provided on the side wall of the casing 1 , and a cooling fan 29 is installed on the side wall of the casing 1 through a cooling fan fixing bracket 30 and is located beside the cooling hole 22 . The power module 3 is installed with the controller 2 through the fixing plate 26 , and the power module 3 is located at the bottom of the fixing plate 26 . The controller 2 is located above the fixed plate 26 . After the power module 3 , the fixing plate 26 and the controller 2 are assembled into a control module, it is installed on the housing 1 through the fixing bracket 25 and the height is adjustable. The indicator light 21 is installed on the controller 2 and exposed on the surface of the housing 1 to indicate whether the power switch 5 is turned on. The spectral sensor 7 is connected to the heat dissipation copper sheet 31 and then electrically connected to the controller 2 . The finger housing 8 is clamped with the housing 1 through a rubber sealing ring 28 . The semiconductor cooling sheet 32 is attached to the heat dissipation copper sheet 31 on one side and the heat dissipation sheet 4 on the other side through the heat-conducting silica gel. A stainless steel heat dissipation etched net 27 is also provided on the housing 1 , and the position of the stainless steel heat dissipation etched net 27 is attached to the heat sink 4 . The material of the etched stainless steel heat dissipation mesh 27 is light and thin, with a thickness of only 0.2-0.3mm, and is easy to install. Compared with the heat dissipation mesh made of plastic, it is more beautiful. At the same time, compared with plastic materials, metal materials have better thermal conductivity, which can improve heat dissipation efficiency. Due to the temperature deviation, for the spectral sensor, the detection value will drift, so the heat dissipation of the spectral sensor is extremely strict. After the power switch 5 is turned on, the semiconductor refrigerating sheet 32 is connected to electricity, and the semiconductor refrigerating sheet 32 will be refrigerated and heated while being energized. Since the heat dissipation copper sheet 31 is in contact with the spectral sensor 7, heat conduction can be realized, thereby ensuring that the ambient temperature is controllable when the spectral sensor 7 detects. Moreover, airflow is formed in the casing 1 through the heat dissipation fan 29, combined with the stainless steel heat dissipation etched mesh 27 for rapid heat dissipation, which can ensure that the detection display of the spectral sensor 7 will not occur due to the ambient temperature when the equipment works for a long time. Value drift.

The non-invasive blood glucose detector with multi-sensory and highly accurate data of the present application has an internal structure as shown in Figures 4 to 7. The LED light source 6 is installed on the housing 1 and electrically connected to the controller 2. At the position of the top surface of the end of the bin 8, the LED light source 6 and the spectral sensor 7 are located on the same straight line. A notch 9 is provided at the end of the finger bin 8 , and a slope block 10 is provided at the edge of the notch 9 . The slope stopper 10 is provided with a plurality of temperature sensors 11 arranged side by side, the temperature sensors 11 are electrically connected to the controller 2, and the slope stopper 10 is provided with a through hole 12 for detecting light passing through.

The non-invasive blood glucose detector also includes a fingertip pressure detection mechanism 13, the fingertip pressure detection mechanism 13 includes a detection bracket 14, a fingertip stopper 15 and a pressure sensor 16, the detection bracket 14 is arranged on the housing 1, the The fingertip block 15 is connected with the detection bracket 14 through the pressure sensor 16 . The pressure sensor 16 is located at the notch 9 and at the edge of the slope block 10 , and the pressure sensor 16 is also electrically connected to the controller 2 . The detection bracket 14 is detachably connected to the housing 1 , and the detection bracket 14 also abuts against the controller 2 . This design can ensure that the position of the detection bracket 14 is stable without deviation. A guide spring 17 is also included, and the guide spring 17 is nested on the pressure sensor 16 . The guide spring 17 is designed with reference to the size and parameters of GB/T1973.3-2005 small cylindrical helical compression spring, the middle diameter of the spring is 0.8mm, and the test load is 2.16N. The guide spring 17 and the pressure sensor 16 are inclined, and the specific angle can be 30 degrees. The fingertip block 15 is provided with a guide rail 19 on the side wall of the finger compartment 8, and the fingertip block 15 slides through the guide rail 19 and the guide rail groove 18. Cooperate. The inner side of the soft material layer of the fingertip stopper 15 is a sponge foam material, and the outer layer is lined with a soft fabric, which cooperates with guide rails and springs, so that when the fingers are inserted, they are not pressed by the components and have a certain damping effect. The position of the fingertip is limited. The fingertip stopper 15 is provided with a fingertip limiting groove 20 with a width of 5 mm to 1 cm; when the fingertip is inserted, the fingertip will automatically keep horizontal and centered according to the tactile habit, and the fingertip stopper 15 shown in the figure The top is arc-shaped, and the height of the notch 9 is greater than that of the fingertip block 15, so that the nail can exceed the fingertip block 15 without being squeezed by the end of the finger bin 8.

The method for specifically detecting the finger, as shown in Figures 3 to 9, includes the following steps:

Step 1, turn on the power switch 5, and the power module 3 powers on the LED light source 6, the spectrum sensor 7, the temperature sensor 11 and the pressure sensor 16 respectively through the controller 2;

Step 2. Insert your finger into the finger compartment 8 and probe into the bottom of the finger compartment 8. The nail part of the fingertip will be located in the gap between the fingertip stopper 15 and the notch 9, so as to ensure that the length of the nail will not match the finger. The detection part of the tip causes interference, the front end of the fingertip is limited by the fingertip limit groove 20 and the fingertip is horizontal, and the guide spring 17 provides a damping effect and limits the forward displacement of the fingertip;

Step 3, when the pressure sensor 16 receives the pressure signal and remains stable, it is determined that the fingertip has reached the bottom of the finger compartment 8 and remains horizontal, and the pressure value received by the pressure sensor 16 is a coefficient K; when the temperature sensor 11 receives the temperature change , thereby judging the fingertip temperature T1, the unchanged temperature is the ambient temperature T2, and the number N of changes in the temperature sensor 11;

Step 4: After the light of the specified wavelength is emitted by the LED light source 6, the light passes through the through hole 12 and then passes through the detection part at the front of the fingertip. The spectral sensor 7 is equipped with a Fabry-Perot interferometer tunable filter, and Adjust the optical receiving range of the tunable filter to nm level and perform spectrum receiving;

Step 5. According to the number N of changes in the temperature sensor 11, determine whether the type of the finger belongs to the first, second or third category, combine the coefficient K, the fingertip temperature T1 and the ambient temperature T2, and combine the spectral sensor 7 to receive light The converted ADC value after the signal is jointly sent to the controller 2, and the controller 2 invokes the algorithm in the built-in data storage module to calculate the blood glucose value. Among them, the fingers are divided into three categories, the first category is the thumb, the second category is the index finger, middle finger and ring finger, and the third category is the little finger.

The LED light source 6 emits specified wavelengths of 1500nm, 1525nm, 1550nm and 1575nm.

In step two, the unique material design of the fingertip block 15 can not only ensure that when the end of the finger has a blocking effect, there will be no excessive extrusion and the movement direction will be moved by the direction of the guide rail 19 and the direction of the guide spring 17 Restricted, basically level and without deflection.

When the power switch 5 is turned on, the power module 3 also energizes the semiconductor cooling chip 32 through the controller 2 . The temperature of the spectral sensor 7 is directly lowered by the semiconductor cooling sheet 32 through the heat dissipation copper sheet 31 , so that the spectral sensor 7 can sense and receive light at a specified temperature.

The algorithm used in this application usually contains some errors based on the spectral information measured by the spectral sensor, such as stray light, the influence of human tissue, etc., which makes the measured data have certain noise and affects the calculation accuracy of blood sugar. Therefore, before modeling It is necessary to preprocess the collected raw spectral data. The preprocessing includes that the type of the finger belongs to the first type, the second type or the third type, combining coefficient K, fingertip temperature T1 and ambient temperature T2, etc., thereby reducing errors and extracting effective information in the data. The calculation accuracy of the blood glucose model is improved through the above preprocessing. According to the support vector machine theory, the multi-mode spectral data is modeled using MATLAB software.

Two sets of near-infrared spectral data with different wavelengths are input as the independent variable matrix of the model, and the blood glucose value adopted by the household blood glucose meter is used as the dependent variable of the model, and the sample training set and test set are divided. In the early stage, a total of 8 sets of data were measured by testing diabetic patients as modeling data. The screened spectral data are sorted according to the blood glucose values measured by the household blood glucose meter, and the training set and the test set are divided according to the ratio of 3:1 to ensure that the selected samples cover all the blood glucose values, that is, 6 sets of sample data are used as training Set, 2 sets of sample data as the test set. Then normalize the sample data. Two sets of spectral data measured at different wavelengths are used as independent variables, and blood glucose values are used as dependent variables. They are normalized respectively, and the probability distribution is in the same range, so as to reduce the impact on the modeling results due to the large range of data distribution and inconsistent order of magnitude, and Improve training efficiency. To set the SVM parameters, it is necessary to select the best kernel function, the penalty factor coefficient (c) and the parameter coefficient (g) of the kernel function. Due to the differences in models and data, it is impossible to obtain the optimal parameters before modeling. In the process, the method of cross-validation is used to obtain the best values of c and g for training and prediction. There are 6 sets of data in the experimental training set. The spectral data and the true value of blood sugar of 6 groups of samples are used for training to obtain the model between the spectral data and the true value of blood sugar. Substituting the two sets of data of the test set into the model for calculation, the calculated value of the blood glucose of the two sets of data can be obtained.

For the established calibration model, four indicators are used to evaluate the model: the correlation coefficient R, the root mean square error of the calibration set (RMSEC), the root mean square error of the test set (RMSEP) and the relative error E, where the correlation coefficient reflects the predicted value The degree of similarity to the theoretical value, the root mean square error and the relative error reflect the model accuracy.

When using two specific wavelength spectral data for modeling, the correlation coefficient of the training set is 97.29%, the root mean square error is 0.3558mmol/L, the correlation coefficient of the test set is 96.3%, the root mean square error is 0.3804mmol/L, the maximum The relative error is 13.68%, and the average relative error is 0.069%.

The above are the preferred embodiments of the present application, and those skilled in the art of the present invention can also change and modify the above-mentioned embodiments. Therefore, the present invention is not limited to the above-mentioned specific embodiments. Any obvious improvements, replacements or modifications made on the basis of the above-mentioned methods belong to the protection scope of the present application.

Claims

A non-invasive blood glucose detector with multi-sensing and highly accurate data, including a housing (1), a controller (2), a power module (3), a heat sink (4), a power switch (5), and an LED light source (6 ) and a spectral sensor (7), the housing (1) is provided with a finger compartment (8), the power module (3) is electrically connected to the controller (2) and is all located in the housing (1), The LED light source (6) is arranged at the end of the finger compartment (8) and is located on the top surface of the finger compartment (8), and the heat sink (4) is connected with a spectral sensor (7), and the spectral sensor (7) is located at The end of the finger bin (8) and the bottom surface of the finger bin (8), the LED light source (6) and the spectral sensor (7) are located on the same straight line,

The end of the finger bin (8) is provided with a notch (9), and the edge of the notch (9) is provided with a slope stopper (10), and the slope stopper (10) is provided with side-by-side A plurality of temperature sensors (11), the temperature sensors (11) are electrically connected to the controller (2), and the slope stopper (10) is provided with a through hole (12) for detecting the passage of light, and also includes A fingertip pressure detection mechanism (13), the fingertip pressure detection mechanism (13) includes a detection bracket (14), a fingertip stopper (15) and a pressure sensor (16), and the detection bracket (14) is arranged on the shell On the body (1), the fingertip block (15) is connected with the detection bracket (14) through a pressure sensor (16); the pressure sensor (16) is located at the notch (9) and is located at the slope block (10 ), the pressure sensor (16) is also electrically connected with the controller (2).
The multi-sensing and highly accurate non-invasive blood glucose detector according to claim 1, characterized in that:

The detection bracket (14) is detachably connected to the housing (1), and the detection bracket (14) also offsets the controller (2).
According to claim 1 or 2, the multi-sensory and highly accurate non-invasive blood glucose detector is characterized in that:

It also includes a guide spring (17), the guide spring (17) is nested on the pressure sensor (16), and the guide spring (17) and the pressure sensor (16) are inclined.
The multi-sensory and highly accurate non-invasive blood glucose detector according to any one of claims 1-3, characterized in that:

The side wall of the finger compartment (8) is provided with a guide rail groove (18), and the fingertip stopper (15) is provided with a guide rail (19), and the fingertip stopper (15) passes through the guide rail (19) and The guide rail groove (18) is sliding fit.
The multi-sensory and highly accurate non-invasive blood glucose detector according to any one of claims 1-4, characterized in that:

The fingertip stopper (15) is provided with a fingertip limiting groove (20) with a width of 5 mm to 1 cm.
The multi-sensory and highly accurate non-invasive blood glucose detector according to any one of claims 1-5, characterized in that:

The top of the shown fingertip stopper (15) is arc-shaped, and the height of the notch (9) is greater than that of the fingertip stopper (15).
The multi-sensory and highly accurate non-invasive blood glucose detector according to any one of claims 1-6, characterized in that:

It also includes a semiconductor cooling sheet (32) and a heat dissipation copper sheet (31). The cooling end of the semiconductor cooling sheet (32) is attached to the heat dissipation copper sheet (31), and the heating end of the semiconductor cooling sheet (32) is attached to the cooling fin (4).
A non-invasive blood glucose detection method with multi-sensing and highly accurate data, characterized in that: comprising the following steps:

Step 1. Turn on the power switch (5), and the power module (3) powers on the LED light source (6), spectrum sensor (7), temperature sensor (11) and pressure sensor (16) respectively through the controller (2);

Step 2. Insert your finger into the finger compartment (8) and probe into the bottom of the finger compartment (8). The nail part of the fingertip will be located in the gap between the fingertip block (15) and the notch (9). The front end of the tip is limited by the fingertip limit groove (20) and makes the fingertip horizontal, and the damping effect is provided by the guide spring (17) and the forward displacement of the fingertip is limited;

Step 3, when the pressure sensor (16) receives the pressure signal and remains stable, then it is determined that the fingertip has reached the bottom of the finger compartment (8) and remains horizontal, and the pressure value received by the pressure sensor (16) is coefficient K; the temperature sensor ( 11) When the received temperature changes, thereby judging the fingertip temperature T1, the unchanged temperature is the ambient temperature T2, and the number N of changes in the temperature sensor (11);

Step 4: After the LED light source (6) emits light of a specified wavelength, the light passes through the through hole (12) and then passes through the detection site at the front of the fingertip. The spectral sensor (7) is equipped with a Fabry-Perot interferometer Tunable filter, and adjust the optical receiving range of the tunable filter to nm level to receive spectrum;

Step 5, according to the number N of changes in the temperature sensor (11), judge that the type of the finger belongs to the first category, the second category or the third category, combine the coefficient K, fingertip temperature T1 and ambient temperature T2, and combine the spectral sensor ( 7) The converted ADC value after receiving the light signal is sent to the controller (2), and the blood sugar value is calculated by calling the algorithm in the built-in data storage module through the controller (2); among them, the fingers are divided into three categories, the first category is The thumb, the second category is the index finger, middle finger and ring finger, and the third category is the little finger.
The multi-sensory and highly accurate non-invasive blood glucose detection method according to claim 8, characterized in that: the LED light source (6) emits specified wavelengths of 1500nm, 1525nm, 1550nm and 1575nm.
According to claim 8 or 9, the multi-sensory and highly accurate non-invasive blood sugar detection method is characterized in that: when the power switch (5) is turned on, the power module (3) is also a semiconductor through the controller (2). The refrigerating sheet (32) is energized, and the semiconductor refrigerating sheet (32) directly cools the spectral sensor (7) through the heat dissipation copper sheet (31), so that the spectral sensor (7) performs light induction reception at a specified temperature.
A blood glucose detection method utilizing the multi-sensory and highly accurate non-invasive blood glucose detector according to any one of claims 1-7, comprising the steps of:

Step 1. Turn on the power switch (5), and the power module (3) powers on the LED light source (6), spectrum sensor (7), temperature sensor (11) and pressure sensor (16) respectively through the controller (2);

Step 2. Insert your finger into the finger compartment (8) and probe into the bottom of the finger compartment (8). The nail part of the fingertip will be located in the gap between the fingertip block (15) and the notch (9). The front end of the tip is limited by the fingertip limit groove (20) and makes the fingertip horizontal, and the damping effect is provided by the guide spring (17) and the forward displacement of the fingertip is limited;

Step 3, when the pressure sensor (16) receives the pressure signal and remains stable, then it is determined that the fingertip has reached the bottom of the finger compartment (8) and remains horizontal, and the pressure value received by the pressure sensor (16) is coefficient K; the temperature sensor ( 11) When the received temperature changes, thereby judging the fingertip temperature T1, the unchanged temperature is the ambient temperature T2, and the number N of changes in the temperature sensor (11);

Step 4: After the LED light source (6) emits light of a specified wavelength, the light passes through the through hole (12) and then passes through the detection site at the front of the fingertip. The spectral sensor (7) is equipped with a Fabry-Perot interferometer Tunable filter, and adjust the optical receiving range of the tunable filter to nm level to receive spectrum;

Step 5, according to the number N of changes in the temperature sensor (11), judge that the type of the finger belongs to the first category, the second category or the third category, combine the coefficient K, fingertip temperature T1 and ambient temperature T2, and combine the spectral sensor ( 7) The converted ADC value after receiving the light signal is sent to the controller (2), and the blood sugar value is calculated by calling the algorithm in the built-in data storage module through the controller (2); among them, the fingers are divided into three categories, the first category is The thumb, the second category is the index finger, middle finger and ring finger, and the third category is the little finger.