Portable non-invasive blood glucose measurement device and method based on multi-dimensional optical information

Abstract

To overcome the technical problems of traditional Mueller matrix optical rotation method blood glucose measurement devices, such as high energy consumption, large size, low measurement accuracy, and poor reliability, this invention proposes a portable non-invasive blood glucose measurement device and method based on multidimensional optical information. The device includes a light source, a detection module, a probe module, and a processing module. The detection module includes an upper cover plate, a polarization unit, a polarization analyzer unit, and a lower cover plate arranged coaxially from top to bottom and connected in sequence. The light source and the detection module are respectively mounted on the lower cover plate, which has a first incident window and a first detection window corresponding to the light source and the detection module, respectively. The upper cover plate has a second incident window and a second detection window corresponding to the light source and the detection module, respectively. The processing module is connected to the detection module, and the detection module is used to transmit the received reflected light signal to the processing module.

Description

Technical Field

[0001] This invention relates to the field of human blood glucose detection, and in particular to a portable non-invasive blood glucose measurement device and method based on multidimensional optical information. Background Technology

[0002] According to statistics from the IDF (The International Diabetes Federation), the number of people with diabetes worldwide is increasing year by year, becoming a typical disease affecting people's lives. Currently, the commonly used method for treating diabetes is to measure blood glucose levels in combination with oral hypoglycemic drugs and insulin injections. High-precision blood glucose measurement is a key aspect of diabetes treatment. Therefore, blood glucose concentration detection plays a crucial role in the prevention, diagnosis, and treatment of diabetes.

[0003] Currently, the most common clinical method for measuring blood glucose concentration is through finger-prick blood sampling combined with electrochemical or photochemical analysis. Frequent invasive measurements can be painful for patients. Therefore, there is an urgent need for a portable, non-invasive method for measuring blood glucose concentration that allows for painless, rapid, and safe detection.

[0004] In recent years, various non-invasive blood glucose measurement methods have been proposed both domestically and internationally. Among them, optical methods such as Raman spectroscopy, infrared spectroscopy, and optical rotation have attracted much attention due to their advantages such as fast measurement speed, non-invasiveness, and multidimensional information. For example, Chinese invention patent CN115177243A discloses a miniature wearable wristwatch-style non-invasive blood glucose monitoring system based on Raman spectroscopy. This system integrates a near-infrared ultra-narrow bandwidth light source with fiber optic output and an avalanche photodiode into a wearable wristwatch to collect Raman signals, and then uses algorithms such as deconvolution to obtain high-precision blood glucose prediction results.

[0005] However, the signal measured using Raman spectroscopy is relatively weak, and the measurement is affected by temperature, causing drift. Infrared spectroscopy has drawbacks such as indistinct glucose absorption peaks and significant individual variability. These factors greatly affect the accuracy and stability of non-invasive blood glucose testing.

[0006] Traditional optical rotation methods measure blood glucose based on the linear relationship between the concentration of glucose in the blood and the optical rotation angle. However, due to the complexity of human blood and tissue composition, it is difficult to achieve high-precision measurement of human blood glucose directly through a single optical rotation angle. By introducing the Mueller matrix, which can completely describe the optical properties of complex and chaotic human tissues, into the optical rotation method, the multidimensional optical parameters in the Mueller matrix optical rotation method can be effectively improved to enhance measurement accuracy.

[0007] Traditional Mueller matrix optical rotation method blood glucose measurement devices typically employ a polarization state controller composed of multiple discrete polarizers, waveplates, and photoelectric modulators, combined with optical components such as a light source, collimating lens, and condenser lens to measure blood glucose concentration. These devices suffer from significant light signal energy loss, are bulky, inconvenient to carry and operate, and expensive, hindering daily blood glucose monitoring for diabetic patients and the market promotion of non-invasive blood glucose meters. Furthermore, traditional optical rotation methods predict blood glucose concentration based on the linear relationship between a single optical variable (optical rotation angle) and blood glucose concentration, resulting in low accuracy, poor reproducibility, and low reliability, making it difficult to meet the demands for high-precision blood glucose measurement in complex real-world human settings. Summary of the Invention

[0008] To overcome the technical problems of traditional Mueller matrix optical rotation method blood glucose measurement devices, such as high energy consumption, large size, low measurement accuracy, and poor reliability, this invention proposes a portable non-invasive blood glucose measurement device and method based on multidimensional optical information.

[0009] To achieve the above objectives, the technical solution of the present invention is as follows:

[0010] A portable non-invasive blood glucose measurement device based on multidimensional optical information is characterized by including a light source, a detection module, a probe module, and a processing module.

[0011] The detection module includes an upper cover plate, a deflection unit, a deflection detection unit, and a lower cover plate, which are coaxially arranged and connected sequentially from top to bottom.

[0012] The light source and the detection module are respectively disposed on the lower cover plate, and the lower cover plate is provided with a first incident window and a first detection window corresponding to the light source and the detection module, respectively; the upper cover plate is provided with a second incident window and a second detection window corresponding to the light source and the detection module, respectively.

[0013] The incident light from the light source passes through the first incident window, the central hole set on the polarization analyzer, the polarization unit, and the second incident window in sequence to reach the human body detection site. After diffuse reflection, it forms reflected light. The reflected light then passes through the second detection window, the third detection window set on the polarization analyzer, the first detection window in sequence and is received by the detection module.

[0014] The processing module is connected to the detection module, and the detection module is used to transmit the received reflected light signal to the processing module.

[0015] Furthermore, the biasing unit includes a first spacer and a first gear disposed within the first spacer; the first gear is connected to the upper cover plate via a cylindrical pin, and the first gear can be rotated via the cylindrical pin.

[0016] The first gear is provided with four first polarization holes arranged in a 2×2 array. In the clockwise direction, the four first polarization holes are respectively provided with a 0° linear polarizer, a 45° linear polarizer, a 90° linear polarizer and a circular polarizer for polarizing the incident light between the second incident window and the central hole on the analyzer unit.

[0017] The detection unit includes an annular second spacer and a second gear coaxially disposed within the second spacer; the second gear is connected to the lower cover plate via a hollow rotating shaft, and the second gear can rotate along the hollow rotating shaft; the central hole on the detection unit is located inside the hollow rotating shaft;

[0018] The second gear is provided with four second polarization holes arranged in a 2×2 array. Each of the four second polarization holes is provided with a 0° linear polarizer, a 45° linear polarizer, a 90° linear polarizer, and a circular polarizer for polarizing the reflected light between the third detection window and the first detection window.

[0019] The upper cover plate, the first spacer, the second spacer, and the lower cover plate are connected in sequence.

[0020] Furthermore, the first spacer and the second spacer are respectively provided with ultrasonic motors connected to the first gear and the second gear, for driving the first gear and the second gear to rotate.

[0021] Furthermore, the first spacer has a first notch on the side edge away from the third detection window for manually rotating the first gear; the second spacer has a second notch for manually rotating the second gear.

[0022] Furthermore, the distance D1 between the central axis of the first incident window and the central axis of the first detection window, and the distance D2 between the central axis of the second incident window and the central axis of the second detection window are equal, and the values ​​of D1 and D2 are 8-12mm.

[0023] Furthermore, a sponge ring is provided on the upper cover plate, and the lower end face of the sponge ring is connected to the edge of the upper cover plate to block ambient light;

[0024] The second incident window and the second detection window are respectively equipped with rubber rings for direct contact with the part of the human body to be tested.

[0025] Furthermore, the outer sides of the first gear and the second gear are respectively provided with dot-shaped markings that correspond one-to-one with each polarizer, which are used to indicate the location of the polarizer.

[0026] This invention also provides a portable non-invasive blood glucose measurement method based on multidimensional optical information, characterized in that it employs the aforementioned portable non-invasive blood glucose measurement device based on multidimensional optical information, and includes the following steps:

[0027] Step 1】Turn on the light source. The incident light passes through the first incident window, the central hole set on the polarization analyzer, the polarization unit, and the second incident window in sequence before reaching the human body detection area. After diffuse reflection by the human body detection area, it forms reflected light.

[0028] Step 2】The reflected light passes sequentially through the second detection window, the third detection window, the polarization analyzer, and the first detection window before entering the detection module;

[0029] Step 3】The detection module transmits the received reflected light signal to the processing module;

[0030] Step 4: The processing module processes the received reflected light signal to obtain non-invasive blood glucose measurement information based on multi-dimensional optical information.

[0031] Further, in step 1], when the incident light passes through the polarizing unit, the first gear is rotated so that the incident light passes through the 0° linear polarizer, 45° linear polarizer, 90° linear polarizer and circular polarizer respectively set in the first polarizing aperture.

[0032] In step 2, when the reflected light passes through the polarization analyzer, the second gear is rotated so that the reflected light passes through the 0° linear polarizer, 45° linear polarizer, 90° linear polarizer and circular polarizer respectively set in the second polarization aperture.

[0033] Furthermore, in step 4], the reflected light signal received by the processing module is a digital signal, which is a set of 4×4 Mueller matrices. The specific processing procedure is as follows:

[0034] 4.1】The Mueller matrix was decomposed and measured by the Mueller matrix polarization decomposition method to obtain 16-dimensional optical parameters; the optical parameters include optical rotation angle, phase retardation, dichroism, degree of polarization, and degree of depolarization;

[0035] 4.2】Effective optical parameters are obtained by calculating the correlation coefficient between optical parameters and standard blood glucose values;

[0036] 4.3】An optical information measurement space is constructed based on effective optical parameters by using a spatial metric learning dimensionality reduction method; the optical information measurement space includes a 1D measurement space, a 2D measurement space, and a 3D measurement space;

[0037] 4.4】By calculating the evaluation function of the optical information metric space, the optimal metric space data information is selected and normalized preprocessed.

[0038] 4.5】The preprocessed information is input into the neural network for training to obtain the trained output information;

[0039] 4.6】The trained output information is denormalized to obtain the blood glucose value.

[0040] Compared with the prior art, the present invention has the following beneficial technical effects:

[0041] 1. The portable non-invasive blood glucose measurement device of the present invention includes a light source, a detection module, a probe module and a processing module. The detection module includes an upper cover plate, a polarization unit, a polarization detection unit and a lower cover plate arranged coaxially from top to bottom and connected in sequence. The light source and the probe module are respectively located in the lower cover plate. The device has a high degree of integration, small size and convenient portability, which is in line with the development trend of wearable, lightweight and high-precision non-invasive blood glucose devices.

[0042] 2. The detection module of the present invention includes a polarization unit and a polarization detection unit, which can realize intelligent and convenient integrated measurement of the optical parameters of the 16-dimensional Mueller matrix, avoiding the problems of serious energy loss and multiple calibrations required by traditional discrete optical systems. It has small measurement error, high accuracy and strong reliability.

[0043] 3. The first spacer and the second spacer of the present invention can be driven by an ultrasonic motor to rotate the first gear and the second gear respectively; or the first gear and the second gear can be rotated manually, thereby realizing the rapid measurement of the 16-dimensional Mueller matrix optical parameters, improving the convenience and efficiency of measurement.

[0044] 4. The upper cover plate of the present invention is provided with a sponge ring, which is used to block ambient light, thereby making the test results more accurate.

[0045] 5. In the blood glucose measurement method of the present invention, an optical information measurement space is constructed based on the obtained 16-dimensional optical parameters through a spatial metric learning dimensionality reduction method. This enables the rapid sorting of effective optical information and redundant information, maximizes the integration and utilization of effective optical information, improves information utilization, and thus improves test accuracy.

[0046] 6. In the blood glucose measurement method of the present invention, the advantages of high-dimensional spatial data information and high-precision prediction model of neural network can be fully utilized through neural network training. Error optimization is achieved by iteratively calculating the neural network big data multiple times, thereby improving the detection accuracy of blood glucose value. Attached Figure Description

[0047] Figure 1 This is a schematic diagram of the structure of an embodiment of the portable non-invasive blood glucose measurement device based on multidimensional optical information of the present invention;

[0048] Figure 2 This is a schematic diagram of the information processing process in an embodiment of the portable non-invasive blood glucose measurement method based on multidimensional optical information of the present invention;

[0049] Figure 3 is a schematic diagram of the neural network prediction results in an embodiment of the portable non-invasive blood glucose measurement method based on multidimensional optical information of the present invention. Figure 3a , Figure 3c , Figure 3e These represent the errors in the 1D original space, 2D original space, and 3D original space, respectively. Figure 3b , Figure 3d , Figure 3f These represent the prediction errors in the 1D, 2D, and 3D metric spaces after metric space learning, respectively.

[0050] The attached figures are labeled as follows:

[0051] 1-Light source, 2-Detection module, 3-Processing module, 4-Upper cover plate, 41-Second incident window, 42-Second detection window, 5-Polarization unit, 51-First spacer, 511-Third detection window, 512-First notch, 52-First gear, 521-First polarization aperture, 6-Polarization analyzer, 61-Second spacer, 611-Second notch, 62-Second gear, 621-Second polarization aperture, 7-Lower cover plate, 71-First incident window, 72-First detection window, 8-Sponge ring, 9-Rubber ring. Detailed Implementation

[0052] To make the objectives, technical solutions, and advantages of this invention clearer, the following detailed description of a portable non-invasive blood glucose measurement device and method based on multidimensional optical information, in conjunction with the accompanying drawings and specific embodiments, is provided. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of this invention and are not intended to limit the scope of protection of this invention.

[0053] like Figure 1 As shown, this embodiment provides a portable non-invasive blood glucose measurement device based on multidimensional optical information, including a light source 1, a detection module, a probe module 2, and a processing module 3.

[0054] The detection module includes an upper cover plate 4, a deflection unit 5, a deflection detection unit 6, and a lower cover plate 7, which are coaxially arranged and connected sequentially from top to bottom. Both the upper cover plate 4 and the lower cover plate 7 are plastic covers, but covers of appropriate materials can also be selected according to specific design requirements.

[0055] The biasing unit 5 includes a first spacer 51 and a first gear 52 disposed in the first spacer 51; the first gear 52 is connected to the upper cover plate 4 by a cylindrical pin, and the first gear 52 can be rotated by the cylindrical pin.

[0056] The first spacer 51 is also provided with a third detection window 511.

[0057] The polarization detection unit 6 includes an annular second spacer 61 and a second gear 62 coaxially disposed within the second spacer 6; the second gear 62 is connected to the lower cover plate 7 via a hollow rotating shaft, and the second gear 62 can rotate along the hollow rotating shaft, which is used to connect the light source.

[0058] The first gear 52 has four first polarization apertures 521 arranged in a 2×2 array. Each of the four first polarization apertures 521 contains a 0° linear polarizer, a 45° linear polarizer, a 90° linear polarizer, and a circular polarizer, respectively, for polarizing the incident light between the second incident window 41 and the central hole on the analyzer unit 6. The second gear 62 has four second polarization apertures 621 arranged in a 2×2 array. Each of the four second polarization apertures 621 contains a 0° linear polarizer, a 45° linear polarizer, a 90° linear polarizer, and a circular polarizer, respectively, for polarizing the reflected light between the third detection window 511 and the first detection window 72. These four types of polarizers are used to generate and detect specific polarization states. During measurement, the intensity of the 4×4 full optical field can be obtained by corresponding to each selected region. The outer sides of the first gear 52 and the second gear 62 are respectively provided with dot-shaped markers corresponding to each polarizer, indicating the location of the polarizer.

[0059] In this embodiment, the first spacer 51 and the second spacer 61 are respectively provided with ultrasonic motors connected to the first gear 52 and the second gear 62, which are used to drive the first gear 52 and the second gear 62 to rotate automatically, so as to achieve high-precision positioning, efficient acquisition of detection data and rapid detection.

[0060] Alternatively, a first notch 512 is provided on the side edge of the first spacer 51 away from the third detection window 511 for manually rotating the first gear 52; a second notch 611 is provided on the second spacer 61 for manually rotating the second gear 62.

[0061] Alternatively, depending on the actual design, an ultrasonic motor may be provided in both the first spacer 51 and the second spacer 61, and notches may be provided on both the first spacer 51 and the second spacer 61, thereby achieving a combination of automatic and manual drive.

[0062] The light source 1 and the detection module 2 are respectively mounted on the lower cover plate 7, and are integrated into the same circuit system. The lower cover plate 7 has a first incident window 71 and a first detection window 72 corresponding to the light source 1 and the detection module 2, respectively. The circuit system also integrates a built-in battery and a trigger. The built-in battery provides power to the light source 1 and the detection module 2, and the trigger is used to turn on the system and trigger the information acquisition function of the detection module 2.

[0063] The upper cover plate 4 is provided with a second incident window 41 and a second detection window 42 corresponding to the light source 1 and the detection module 2, respectively. Rubber rings 9 are provided on the second incident window 41 and the second detection window 42. During detection, the rubber rings 9 are tightly attached to the part of the human body to be tested, blocking the light from the light source 1 from directly entering the second detection window 42. A sponge ring 8 is also provided on the upper cover plate 4. The lower end face of the sponge ring 8 is connected to the edge of the upper cover plate 4 to block ambient light.

[0064] In this embodiment, the upper cover plate 4, the first spacer 51, the second spacer 61, and the lower cover plate 7 are connected in sequence. The light source 1 is a single light source, which can be an LED or LD with wavelengths of 632nm, 850nm, 1330nm, or 1550nm. A single light source avoids the differences between light sources of different specifications or from the same batch, making the test results more repeatable. The power of the single light source limits the distance between the incident window and the detection window, specifically the distance D1 between the central axes of the first incident window 71 and the first detection window 72, and the distance D2 between the central axes of the second incident window 41 and the second detection window 42. To ensure that the detection window receives sufficient light, D1 and D2 are equal and range from 8 to 12 mm.

[0065] The working principle of this device is as follows: the collimated and parallel incident light emitted by the light source 1 passes sequentially through the first incident window 71, the central hole set on the analyzer unit 6, any polarizer set in the first polarization aperture 521, and the second incident window 41 before reaching the human body detection site. The incident light undergoes diffuse reflection in the subcutaneous tissue to form reflected light. The reflected light then passes sequentially through the second detection window 42, the third detection window 511, any polarizer set in the second polarization aperture 621, and the first detection window 72 before being received by the detection module 2. The detection module 2 is used to convert the received reflected light signal into a digital signal.

[0066] The processing module 3 is connected to the detection module 2 via a data port. The acquisition program of the detection module 2 is set to acquire data twice per test, and the difference in light intensity between the two data acquisitions is calculated. If the difference between two consecutive data acquisitions exceeds a certain range, it indicates an abnormality in the measurement, thus determining the usability of the data. Simultaneously, a manual rotation can be used to issue an acquisition command. After receiving the data, the module can then withdraw, display, and save it. The processing module 3 is used to perform calculations and processing on the digital signal to ultimately obtain the human blood glucose level. In this embodiment, the detection module 2 can also be used with lenses of different focal lengths to meet detection requirements with different signal-to-noise ratios.

[0067] Based on this, this embodiment also provides a portable non-invasive blood glucose measurement method based on multidimensional optical information, including the following steps:

[0068] Step 1: Turn on the light source 1. The incident light passes sequentially through the first incident window 71, the central hole on the analyzer unit 6, the polarizing unit 5, and the second incident window 41 before reaching the human body detection area. After diffuse reflection at the human body detection area, reflected light is formed. In actual measurement, when the incident light passes through the polarizing unit 5, the first gear 52 is rotated so that the incident light passes through the 0° linear polarizer, 45° linear polarizer, 90° linear polarizer, and circular polarizer respectively set in the first polarizing hole 521.

[0069] Step 2】The reflected light sequentially passes through the second detection window 42, the third detection window 511, the polarization analyzer 6, and the first detection window 72 before entering the detection module 2. During actual measurement, as the reflected light passes through the polarization analyzer 6, the second gear 62 is rotated, causing the reflected light to pass through a 0° linear polarizer, a 45° linear polarizer, a 90° linear polarizer, and a circular polarizer located in the second polarization aperture 621. Combining these with the four polarization modes from Step 1】, a set of 4×4 Mueller matrix reflected light signals can be obtained.

[0070] Step 3】Detection module 2 converts the received reflected light signal into a digital signal and transmits it to processing module 3.

[0071] Step 4】The receiving digital signal is optimized by processing module 3, combined with Figure 2 As shown, the optimization process of processing module 3 is as follows:

[0072] 4.1 The Mueller matrix was decomposed and measured by the Mueller Matrix Polar Decomposition (MMPD) method to obtain 16-dimensional optical parameters, including optical rotation angle, phase delay, dichroism, degree of polarization, degree of depolarization, etc.

[0073] 4.2】Effective optical parameters are obtained by calculating the correlation coefficient between optical parameters and standard blood glucose values.

[0074] 4.3】Optical information measurement space is constructed based on effective optical parameters through spatial metric learning dimensionality reduction methods. The optical information measurement space is constructed from effective optical parameters using dimensionality reduction methods such as PCA, MDS, and Factor analysis. The optical information measurement space includes 1D measurement space, 2D measurement space, and 3D measurement space.

[0075] 4.4 By calculating the evaluation functions of 1D, 2D, and 3D metric spaces, and comparing the correlation coefficient of the 1D metric space, and the feature degree and clustering of the 2D and 3D metric spaces, the optimal metric space data information is selected. This data is then normalized to a unit space to obtain the input information for neural network training, thereby improving the prediction speed of the neural network. Alternatively, the normalized preprocessed information can be split into training and testing sets for rapid and reliable subsequent testing of the prediction accuracy of the neural network prediction model.

[0076] In this embodiment, the evaluation functions for metric learning include correlation coefficient, clustering degree, and feature degree. The evaluation function for the 1D metric space is the correlation coefficient, which represents the linearity between two parameters. The evaluation functions for the 2D and 3D metric spaces are clustering degree and feature degree, where clustering degree represents the similarity of glucose concentrations in a group, and feature degree represents the distinctiveness between different groups of glucose concentrations. The specific calculation methods can be based on commonly used evaluation function calculation formulas.

[0077] 4.5】Construct a neural network and set configuration parameters to obtain a neural network model. In this embodiment, the back propagation (BP) neural network model is used as an example. The neural network model includes an input layer and an output layer. The configuration parameters include the maximum number of iterations and the learning rate of the neural network training. In the actual testing process, the above structural parameters of the neural network prediction model can be iteratively optimized according to the test results and test requirements to ensure the accuracy of the final test data.

[0078] 4.6】The input information from step 4.4】is input into the input layer of the neural network model in step 4.5】for training, and the output information is obtained in the output layer.

[0079] 4.7】The output information is denormalized to obtain the blood glucose value.

[0080] Refer to Figure 3 Figure 3a , Figure 3b , Figure 3c , Figure 3d , Figure 3e , Figure 3f It can be seen that among the errors of the 1D original space, 2D original space, and 3D original space and the prediction errors of the 1D metric space, 2D metric space, and 3D metric space after metric space learning, the maximum error of the blood glucose value measured by the 3D original space is within 10% compared with the 1D metric space, 2D metric space, and 3D metric space, which all meet the testing requirements.

[0081] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A portable, non-invasive blood glucose measurement device based on multidimensional optical information, characterized in that: It includes a light source (1), a detection module, a probe module (2), and a processing module (3); The detection module includes an upper cover plate (4), a deflection unit (5), a deflection detection unit (6), and a lower cover plate (7) that are coaxially arranged and connected in sequence from top to bottom. The light source (1) and the detection module (2) are respectively disposed on the lower cover plate (7). The lower cover plate (7) is provided with a first incident window (71) and a first detection window (72) corresponding to the light source (1) and the detection module (2) respectively. The upper cover plate (4) is provided with a second incident window (41) and a second detection window (42) corresponding to the light source (1) and the detection module (2) respectively. The incident light from the light source (1) passes through the first incident window (71), the central hole set on the polarization analyzer (6), the polarization unit (5), and the second incident window (41) in sequence. After diffuse reflection, the reflected light is formed at the human body detection site. The reflected light then passes through the second detection window (42), the third detection window (511) set on the polarization unit (5), the polarization analyzer (6), and the first detection window (72) in sequence before being received by the detection module (2). The polarizing unit (5) includes a first spacer (51) and a first gear (52) disposed in the first spacer (51); the first gear (52) is connected to the upper cover plate (4) by a cylindrical pin, and the first gear (52) can be rotated by the cylindrical pin; the first gear (52) is provided with four first polarizing holes (521) arranged in a 2×2 array, and in the clockwise direction, the four first polarizing holes (521) are respectively provided with a 0° linear polarizer, a 45° linear polarizer, a 90° linear polarizer and a circular polarizer for polarizing the incident light between the second incident window (41) and the center hole on the polarizing unit (6); The polarization detection unit (6) includes an annular second spacer (61) and a second gear (62) coaxially disposed within the second spacer (6); the second gear (62) is connected to the lower cover plate (7) via a hollow rotating shaft, and the second gear (62) can rotate along the hollow rotating shaft; the central hole on the polarization detection unit (6) is located inside the hollow rotating shaft; the second gear (62) is provided with four second polarization holes (621) arranged in a 2×2 array, and the four second polarization holes (621) are respectively provided with holes for the third detection window (511) and the first detection window. Between the windows (72) are 0° linear polarizers, 45° linear polarizers, 90° linear polarizers and circular polarizers for polarizing the reflected light; the upper cover plate (4), the first spacer (51), the second spacer (61) and the lower cover plate (7) are connected in sequence; the upper cover plate (4) is provided with a sponge ring (8), the lower end face of the sponge ring (8) is connected to the edge of the upper cover plate (4) to block ambient light; the processing module (3) is connected to the detection module (2), and the detection module (2) is used to transmit the received reflected light signal to the processing module (3).

2. The portable non-invasive blood glucose measurement device based on multidimensional optical information according to claim 1, characterized in that: The first spacer (51) and the second spacer (61) are respectively provided with ultrasonic motors connected to the first gear (52) and the second gear (62) to drive the first gear (52) and the second gear (62) to rotate.

3. The portable non-invasive blood glucose measurement device based on multidimensional optical information according to claim 1, characterized in that: The first spacer (51) has a first notch (512) on the side away from the third detection window (511) for manually rotating the first gear (52); the second spacer (61) has a second notch (611) for manually rotating the second gear (62).

4. The portable non-invasive blood glucose measurement device based on multidimensional optical information according to claim 2 or 3, characterized in that: The distance D1 between the central axis of the first incident window (71) and the central axis of the first detection window (72) and the distance D2 between the central axis of the second incident window (41) and the central axis of the second detection window (42) are equal, and the values ​​of D1 and D2 are 8-12mm.

5. The portable non-invasive blood glucose measurement device based on multidimensional optical information according to claim 4, characterized in that: Rubber rings (9) are provided on the second incident window (41) and the second detection window (42) respectively, for direct contact with the part of the human body to be tested.

6. The portable non-invasive blood glucose measurement device based on multidimensional optical information according to claim 5, characterized in that: The outer sides of the first gear (52) and the second gear (62) are respectively provided with dot-shaped marks corresponding to each polarizer, which are used to indicate the location of the polarizer.

7. A portable, non-invasive blood glucose measurement method based on multidimensional optical information, characterized in that, The portable non-invasive blood glucose measurement device based on multidimensional optical information as described in any one of claims 1-6 includes the following steps: Step 1】Turn on the light source (1). The incident light passes through the first incident window (71), the central hole set on the polarization analyzer (6), the polarization unit (5), and the second incident window (41) in sequence before reaching the human body detection area. After diffuse reflection by the human body detection area, it forms reflected light. Step 2】The reflected light passes through the second detection window (42), the third detection window (511), the polarization analyzer (6), and the first detection window (72) in sequence before entering the detection module (2); Step 3】The detection module (2) transmits the received reflected light signal to the processing module (3); Step 4】The receiving reflected light signal is processed by the processing module (3) to obtain non-invasive blood glucose measurement information based on multidimensional optical information.

8. The portable non-invasive blood glucose measurement method based on multidimensional optical information according to claim 7, characterized in that: In step 1, when the incident light passes through the polarizing unit (5), the first gear (52) is rotated so that the incident light passes through the 0° linear polarizer, 45° linear polarizer, 90° linear polarizer and circular polarizer respectively set in the first polarizing hole (521). In step 2, when the reflected light passes through the polarization analyzer (6), the second gear (62) is rotated so that the reflected light passes through the 0° linear polarizer, 45° linear polarizer, 90° linear polarizer and circular polarizer respectively set in the second polarization aperture (621).

9. The portable non-invasive blood glucose measurement method based on multidimensional optical information according to claim 8, characterized in that: In step 4], the reflected light signal received by the processing module (3) is a digital signal, which is a set of 4×4 Mueller matrices. The specific processing procedure is as follows: 4.1】The Mueller matrix was decomposed and measured by the Mueller matrix polarization decomposition method to obtain 16-dimensional optical parameters; the optical parameters include optical rotation angle, phase retardation, dichroism, degree of polarization, and degree of depolarization; 4.2】Effective optical parameters are obtained by calculating the correlation coefficient between optical parameters and standard blood glucose values; 4.3】Optical information measurement space is constructed based on effective optical parameters by using spatial metric learning dimensionality reduction methods; The optical information measurement space includes a 1D measurement space, a 2D measurement space, and a 3D measurement space. 4.4】By calculating the evaluation function of the optical information metric space, the optimal metric space data information is selected and normalized preprocessed. 4.5】The preprocessed information is input into the neural network for training to obtain the trained output information; 4.6】The trained output information is denormalized to obtain the blood glucose value.

Images