A non-invasive blood glucose screening system and method combining end-side reasoning and host computer intelligent analysis

Abstract

The present application relates to a kind of end side reasoning and host computer intelligent analysis combination's noninvasive blood glucose screening system and method, belong to biomedical detection technical field.The system includes finger clamp type acquisition device, PPG optical sensor module, main control processing module, communication module and host computer analysis module.PPG optical sensor module emits red light and infrared light to finger tissue and collects photoelectric volume pulse wave signal;Main control processing module carries out pre-processing, quality evaluation and feature extraction to PPG signal, and obtains blood glucose estimation value based on end side blood glucose estimation model reasoning calculation;Communication module sends data to host computer, and host computer carries out trend analysis and risk assessment to data.The present application solves the problem that real-time is poor in prior art due to dependence on cloud or host computer calculation, improves detection efficiency and reduces data transmission burden, and is suitable for noninvasive blood glucose screening and health management.

Description

Technical Field

[0001] This invention relates to the field of biomedical detection technology, and in particular to a non-invasive blood glucose screening system and method that combines end-side inference based on photoplethysmography pulse wave signals with intelligent analysis by a host computer. Background Technology

[0002] Diabetes and abnormal glucose metabolism have become a significant public health issue. Current blood glucose testing methods largely rely on blood sampling, which is invasive and inconvenient. Photoplethysmography (PPG)-based non-invasive testing technology offers advantages such as being non-invasive and allowing for continuous monitoring; however, existing technologies generally depend on host computers or cloud-based systems for data processing, resulting in poor real-time performance, heavy data transmission burdens, and insufficient independent operation capabilities on the device itself. Summary of the Invention

[0003] (a) Purpose of the invention The purpose of this invention is to provide a non-invasive blood glucose screening system and method that combines edge-side inference with host computer intelligent analysis, so as to solve the problems of high dependence on cloud or high-performance host computer computing and insufficient real-time processing capability of the device in the prior art.

[0004] (II) Technical Solution To achieve the above objectives, the present invention provides the following technical solution: A non-invasive blood glucose screening system that combines edge-side inference with host computer intelligent analysis includes a finger clip acquisition device, a PPG optical sensor module, a main control processing module, a communication module, and a host computer analysis module.

[0005] The PPG optical sensor module is used to emit red and infrared light to finger tissue and collect PPG signals; the main control processing module is used for signal preprocessing, signal quality assessment, feature extraction and end-side model inference; the communication module is used for data transmission of detection results and related physiological parameters; and the host computer analysis module is used for data storage, trend analysis and risk assessment.

[0006] This invention also provides a non-invasive blood glucose screening method that combines end-side inference with host computer intelligent analysis, including PPG signal acquisition, signal preprocessing, signal quality assessment, feature extraction, end-side model inference, and host computer analysis steps.

[0007] Preferably, the PPG optical sensor module includes a red light emitting unit, an infrared light emitting unit, and a photoelectric receiving unit for acquiring dual-wavelength PPG signals; wherein the center wavelength of the red light is approximately 660 nm, and the center wavelength of the infrared light is approximately 880 nm.

[0008] Preferably, the main control processing module includes a signal preprocessing unit, a signal quality assessment unit, and a feature extraction unit, used to complete the processing of PPG signals and model inference on the edge side.

[0009] Preferably, the signal quality assessment includes at least one of pulse wave amplitude detection, peak-valley detection success rate judgment, heart rate stability judgment, and motion artifact detection, in order to identify unqualified signals before model calculation.

[0010] Preferably, the feature extraction includes at least two types of features: time-domain features, frequency-domain features, and morphological features. The extracted features are input into a preset blood glucose estimation model for inference calculation.

[0011] Preferably, the communication module is a wireless communication module, and the wireless communication module includes at least one of Bluetooth communication modules.

[0012] Preferably, the host computer analysis module performs historical data management, trend analysis, and risk assessment on the received blood glucose estimation values.

[0013] (III) Beneficial Effects 1. By performing model inference calculations on the device side, the system's real-time performance can be improved and its dependence on external computing resources can be reduced; 2. By transmitting only the estimated blood glucose level and related parameters, the amount of data transmitted can be reduced; 3. By performing edge preprocessing and quality assessment, the fluctuations in results caused by low-quality signals entering the model can be reduced; 4. The host computer can perform trend analysis and risk assessment on historical data to facilitate subsequent health screening and management. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the system structure of the present invention; Figure 2 This is a schematic diagram of the signal processing flow of the present invention; Figure 3 This is a schematic diagram of the device structure of the present invention.

[0015] 1. Finger clip-on data acquisition device; 2. PPG optical sensor module; 3. Main control processing module; 4. Communication module; 5. Host computer analysis module; 11. Upper housing; 12. Lower housing; 13. Red light emitting unit; 14. Infrared light emitting unit; 15. Photoelectric receiving unit; 16. Finger receiving slot; 17. Elastic clamping part; 18. Main control circuit board; 19. Power module. Detailed Implementation

[0016] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. However, it should be understood that these embodiments are only for illustrating the present invention and are not intended to limit the scope of protection of the present invention. Example 1

[0017] like Figure 1 As shown, this embodiment provides a non-invasive blood glucose screening system combining edge-side inference and host computer intelligent analysis, including a finger clip acquisition device 1, a PPG optical sensor module 2, a main control processing module 3, a communication module 4, and a host computer analysis module 5. The finger clip acquisition device 1 is used to fix the finger to be tested and provide a relatively stable optical detection environment. The PPG optical sensor module 2 is configured in conjunction with the finger clip acquisition device 1 to emit red and infrared light to the finger tissue and collect photoplethysmography (PPG) wave signals caused by changes in blood volume. The main control processing module 3 is electrically connected to the PPG optical sensor module 2 and is used to preprocess the PPG signal, assess signal quality, and extract features, and perform edge-side inference calculations based on a preset blood glucose estimation model. The communication module 4 is electrically connected to the main control processing module 3 and is used to send the estimated blood glucose value and related physiological parameters to the host computer analysis module 5. The host computer analysis module 5 is used to store, analyze trends, and perform risk assessment processing on the estimated blood glucose value.

[0018] The center wavelength of the red light can be set to approximately 660 nm, and the center wavelength of the infrared light can be set to approximately 880 nm. The photoelectric receiving unit receives the transmitted or reflected light signal and converts it into an electrical signal to form a dual-wavelength PPG signal. Example 2

[0019] like Figure 2 As shown, this embodiment provides a non-invasive blood glucose screening method combining end-side inference and upper-computer intelligent analysis, including the following steps: S1, emitting red and infrared light to finger tissue through a PPG optical sensor module and collecting photoplethysmography (PPG) wave signals from the finger tissue; S2, preprocessing the PPG signal, including DC drift removal, bandpass filtering, and noise suppression; S3, evaluating the signal quality of the preprocessed signal; S4, extracting features from the signal when the signal quality meets preset conditions; S5, inputting the extracted features into an end-side blood glucose estimation model for inference calculation to obtain a blood glucose estimation value; S6, sending the blood glucose estimation value to the upper-computer for trend analysis and risk assessment via a communication module.

[0020] The signal quality assessment may include at least one of pulse wave amplitude detection, peak-valley detection success rate judgment, heart rate stability judgment, and motion artifact detection; the feature extraction may include at least two of time-domain features, frequency-domain features, and morphological features; the blood glucose estimation model may be a linear regression model, a random forest model, or a neural network model. Example 3

[0021] like Figure 3 As shown, the finger clip acquisition device 1 may include an upper housing 11, a lower housing 12, a red light emitting unit 13, an infrared light emitting unit 14 and a photoelectric receiving unit 15 disposed on the optical path. A finger receiving groove 16 is formed inside the lower housing 12. The finger clip acquisition device 1 may also include an elastic clamping part 17 for clamping the finger. The main control circuit board 18 and the power module 19 are disposed inside the housing or connected to the housing, and are used to realize sensor driving, signal sampling, end-side calculation and power supply.

[0022] In actual use, after the user places their finger into the finger receiving slot 16, the red light emitting unit 13 and the infrared light emitting unit 14 emit light alternately. The photoelectric receiving unit 15 acquires the dual-wavelength PPG signal. After the main control processing module 3 completes the end-side processing, the communication module 4 sends the result to the host computer analysis module 5 for further analysis.

Claims

1. A non-invasive blood glucose screening system combining edge-side inference and host computer intelligent analysis, characterized in that, include: A finger clip-type acquisition device is used to fix the finger to be tested and provide a stable optical detection environment; The PPG optical sensor module, in conjunction with the finger clip-type acquisition device, is used to emit red and infrared light to the finger tissue and acquire photoplethysmography (PPG) signals. The main control processing module is electrically connected to the PPG optical sensor module and is used to perform signal preprocessing, signal quality assessment and feature extraction on the PPG signal, and to perform end-side inference calculation based on a preset blood glucose estimation model to obtain the blood glucose estimation value. The communication module is electrically connected to the main control processing module and is used to send the blood glucose estimate and related physiological parameters to the host computer. The host computer analysis module is connected to the communication module and is used to store, analyze trends, and assess risks of the estimated blood glucose values.

2. The system according to claim 1, characterized in that, The PPG optical sensor module includes a red light emitting unit, an infrared light emitting unit, and a photoelectric receiving unit, used to collect dual-wavelength PPG signals, wherein the center wavelength of the red light is approximately 660nm and the center wavelength of the infrared light is approximately 880nm.

3. The system according to claim 1, characterized in that, The main control processing module includes a signal preprocessing unit, a signal quality assessment unit, and a feature extraction unit.

4. The system according to claim 3, characterized in that, The signal quality assessment includes at least one of pulse wave amplitude detection, peak-valley detection success rate assessment, heart rate stability assessment, and motion artifact detection.

5. The system according to claim 3, characterized in that, The feature extraction includes at least two of the following: time-domain features, frequency-domain features, and morphological features.

6. The system according to claim 1, characterized in that, The communication module is a wireless communication module, which includes at least one of Bluetooth communication modules.

7. The system according to claim 1, characterized in that, The blood glucose estimation model is a supervised learning model based on PPG feature input.

8. A non-invasive blood glucose screening method combining end-side reasoning and host computer intelligent analysis, characterized in that, Includes the following steps: S1: The PPG optical sensor module emits red and infrared light to the finger tissue and collects the photoplethysmography (PPG) signal of the finger tissue; S2: Preprocess the PPG signal, including DC drift removal, bandpass filtering, and noise suppression; S3: Perform signal quality assessment on the preprocessed signal; S4: When the signal quality meets the preset conditions, perform feature extraction on the signal; S5: Input the extracted features into the end-side blood glucose estimation model for inference calculation to obtain the blood glucose estimation value; S6: The estimated blood glucose value is sent to the host computer via the communication module for trend analysis and risk assessment.

9. The method according to claim 8, characterized in that, The blood glucose estimation model is a linear regression model, a random forest model, or a neural network model.

10. The method according to claim 8, characterized in that, The host computer generates blood glucose change trends based on historical blood glucose data and outputs risk warning information.

Images