CN-122004850-A - Noninvasive insulin concentration detection method and device

Abstract

The invention provides a noninvasive detection method and device for insulin concentration, which are used for detecting a surface enhanced Raman scattering spectrum signal capable of reflecting the specific vibration characteristics of insulin molecules by using near infrared light, and simultaneously combining the dielectric constant of subcutaneous tissue liquid and a physiological state to cooperatively judge the insulin concentration in blood, so that the problems of inaccurate measurement result and low sensitivity caused by weak and easily interfered Raman spectrum detection signals excited by infrared light are effectively solved. And by introducing the physiological state as a reference, the method and the device can dynamically eliminate the interference caused by sweating, body temperature fluctuation or motion artifact and the like, so that the insulin concentration and blood sugar concentration detection value with high accuracy and high robustness can be provided under a complex physiological environment.

Inventors

• CHEN WU

Assignees

• 上海医英汇医疗科技有限公司

Dates

Publication Date

20260512

Application Date

20260303

Claims (14)

1. 1. A method for noninvasive detection of insulin concentration comprising: Acquiring a dielectric constant, a physiological state and a surface enhanced Raman scattering spectrum signal which can reflect the specific vibration characteristics of insulin molecules of subcutaneous tissue fluid of a user, wherein the dielectric constant, the physiological state and the surface enhanced Raman scattering spectrum signal are measured by a sensing module attached to the skin surface of the user, and the surface enhanced Raman scattering spectrum signal is obtained under the excitation of a near infrared light source; And cooperatively determining the insulin concentration in the subcutaneous tissue fluid of the user according to the dielectric constant of the subcutaneous tissue fluid, the physiological state and the surface enhanced Raman scattering spectrum signal, wherein the physiological state comprises at least one of rest, movement and stress, and the physiological state is determined according to at least one of heart rate, heart rate variability, blood flow perfusion index, skin conductivity and body surface temperature.
2. 2. The noninvasive insulin concentration detection method according to claim 1, wherein the method for cooperatively determining the insulin concentration in the subcutaneous tissue fluid of the user according to the dielectric constant, the physiological state and the surface-enhanced raman scattering spectrum signal of the subcutaneous tissue fluid comprises the steps of inputting the dielectric constant, the physiological state and the surface-enhanced raman scattering spectrum signal of the subcutaneous tissue fluid of the user as input characteristics into a pre-trained multi-mode fusion calculation model, and reasoning to obtain the insulin concentration in the subcutaneous tissue fluid of the user through the multi-mode fusion calculation model.
3. 3. The method for non-invasive detection of insulin concentration according to claim 2, further comprising: Acquiring a near infrared spectrum signal measured by a sensing module attached to the skin surface of a user; determining the blood glucose concentration in subcutaneous tissue fluid of a user according to the near infrared spectrum signals, wherein the determining method of the blood glucose concentration comprises the steps that the input characteristics of the multi-mode fusion calculation model further comprise near infrared spectrum signals, and the results obtained by reasoning the multi-mode fusion calculation model further comprise the blood glucose concentration in the subcutaneous tissue fluid of the user; And fitting to obtain the variation trend of the ratio based on the ratio of the insulin concentration to the blood glucose concentration in subcutaneous tissue fluid at each moment, and determining the insulin resistance level of the user according to the variation trend.
4. 4. The method of noninvasive detection of insulin concentration according to claim 3, further comprising the step of compensating the time difference between the tissue fluid and the insulin and/or blood glucose concentration in blood in real time according to the blood flow perfusion index by the multi-modal fusion calculation model to obtain the insulin concentration and/or blood glucose concentration in blood.
5. 5. The method for noninvasive detection of insulin concentration according to claim 4, wherein the first calibration is performed on the insulin concentration and/or the blood glucose concentration in the blood, and wherein the method for first calibration comprises correcting the insulin concentration and/or the blood glucose concentration in the blood output by the multi-modal fusion calculation model based on the reference insulin concentration value and/or the reference blood glucose concentration value at a plurality of reference moments input by a user.
6. 6. The method for noninvasive detection of insulin concentration according to claim 3, wherein the second calibration is performed on insulin concentration and/or blood glucose concentration in blood, wherein the second calibration method comprises generating a corresponding correction factor based on systematic characteristic deviation between a multi-modal sensing signal at a time of day reference and a pre-stored multi-modal baseline signal, and systematically correcting the insulin concentration and/or blood glucose concentration in blood output by the multi-modal fusion calculation model in the day according to the correction factor.
7. 7. The method for noninvasive detection of insulin concentration according to claim 3, further comprising the steps of using a pre-trained quality control model to evaluate physical quality of each signal measured by each sensing module in real time to obtain a first confidence coefficient, wherein the evaluation criteria of the physical quality comprise at least one of signal to noise ratio and stability of the signal, using the multi-mode fusion calculation model to evaluate matching degree between an inference result and preset physiological logic in real time to obtain a second confidence coefficient, and outputting the inference result of the multi-mode fusion calculation model when the first confidence coefficient and the second confidence coefficient meet respective preset threshold values simultaneously.
8. 8. The noninvasive insulin concentration detection device is characterized by comprising a flexible housing, a flexible battery, a flexible substrate and a flexible adhesive layer which are arranged in a stacked manner from top to bottom, The flexible adhesive layer is used for being attached to the skin surface of a user, and a plurality of first windows are arranged at corresponding positions on the flexible adhesive layer for the sensing module on the flexible substrate to detect the skin of the user; The flexible substrate is stacked above the flexible adhesive layer, a signal acquisition module and an insulin concentration generation module are arranged on the flexible adhesive layer, the signal acquisition module comprises a first sensing module, a second sensing module and a third sensing module, the first sensing module is used for acquiring the dielectric constant of subcutaneous tissue fluid of a user, which is measured by the sensing module attached to the surface of the skin of the user, the second sensing module is used for acquiring the physiological state of the user, the physiological state comprises at least one of rest, movement and stress, the physiological state is determined according to at least one of heart rate, heart rate variability, blood flow perfusion index, skin conductivity and body surface temperature, the third sensing module is used for acquiring a surface enhanced Raman scattering spectrum signal of the user, the surface enhanced Raman scattering spectrum signal is obtained under the excitation of a near infrared light source, and the insulin concentration generation module is used for cooperatively determining the insulin concentration in the subcutaneous tissue fluid of the user according to the dielectric constant of the subcutaneous tissue fluid, the physiological state and the surface enhanced Raman scattering spectrum signal. The flexible battery is arranged above the flexible substrate to supply power to the flexible substrate and shield interference of external environment on circuit signals on the flexible substrate; the flexible cover shell is covered above the flexible battery, and plays a role in protecting the device.
9. 9. The noninvasive insulin concentration detection apparatus of claim 8, wherein the signal acquisition module further comprises a fourth sensor module, the fourth sensor module is configured to acquire near infrared spectrum signals measured by the sensor module attached to the skin surface of the user, the third sensor module and the fourth sensor module are disposed at the center of the flexible substrate, and the first sensor module is circumferentially distributed along the center of the flexible substrate.
10. 10. The noninvasive insulin concentration sensing device of claim 8, wherein the second sensing module comprises a skin conductivity sensing unit comprising two oppositely disposed electrodes.
11. 11. The noninvasive insulin concentration detection apparatus of claim 10, wherein the third sensor module is provided with a near infrared light source, and the near infrared light source in the third sensor module is turned off when the fifth sensor module included in the skin conductivity detection unit or the signal acquisition module on the second sensor module detects that the flexible adhesive layer is not tightly adhered to the skin.
12. 12. The noninvasive insulin concentration detection apparatus according to claim 8, wherein a second window is provided in a third sensor module area of the flexible substrate, and when the flexible substrate is laminated with the flexible adhesive layer, the second window coincides with a first window so that the near infrared light source irradiates the skin through the second window and the first window, and a raman enhancement substrate is provided on a side of the second window facing the skin, and the raman enhancement substrate is detachably provided on the flexible substrate.
13. 13. The non-invasive insulin concentration detection apparatus according to claim 8, wherein the third sensing module comprises a raman signal acquisition unit, and a flexible metal shielding foil is covered above the raman signal acquisition unit.
14. 14. An electronic terminal, comprising: And a processor coupled to the memory, the processor configured to perform the non-invasive detection of insulin concentration method of any of claims 1-7 based on instructions stored in the memory.

Description

Noninvasive insulin concentration detection method and device Technical Field The disclosure relates to the field of insulin detection, in particular to a noninvasive detection method and device for insulin concentration. Background The existing noninvasive insulin detection method often uses the principle that protein molecules such as insulin have strong resonance Raman response in the ultraviolet spectrum, and uses ultraviolet light to excite Raman spectrum to obtain insulin concentration. However, ultraviolet light has extremely shallow penetration depth, is difficult to reach subcutaneous tissue fluid, and more serious, long-term ultraviolet radiation has potential phototoxicity and cancerogenic risk to human skin, so that the method is not suitable for continuous monitoring of wearable equipment. And the infrared light is adopted to excite the Raman spectrum, so that the Raman spectrum excited by the infrared light has weak signals and noisy background, and the sensor adopting a single physical principle is difficult to specifically identify insulin molecules, so that the measurement accuracy and the sensitivity are low. Moreover, the device is easy to generate larger measurement errors under specific wearing conditions (such as improper wearing, severe exercise and the like), and has poor robustness. Disclosure of Invention In view of the above-mentioned drawbacks of the prior art, an object of the present disclosure is to provide a non-invasive detection method and device for insulin concentration, which are used for solving the aforementioned problems. A first aspect of the present disclosure provides a non-invasive detection method of insulin concentration, comprising acquiring a dielectric constant, a physiological state of subcutaneous tissue fluid of a user, and a surface enhanced Raman scattering spectrum signal capable of reflecting specific vibration characteristics of insulin molecules, which are measured by a sensing module attached to the skin surface of the user, wherein the surface enhanced Raman scattering spectrum signal is obtained under excitation of a near infrared light source, cooperatively determining insulin concentration in the subcutaneous tissue fluid of the user according to the subcutaneous tissue fluid dielectric constant, the physiological state and the surface enhanced Raman scattering spectrum signal, wherein the physiological state comprises at least one of rest, movement and stress, and the physiological state is determined according to at least one of heart rate, heart rate variability, blood perfusion index, skin conductivity and body surface temperature. In an embodiment of the first aspect of the disclosure, the method for cooperatively determining the insulin concentration in the subcutaneous tissue fluid of the user according to the permittivity, the physiological state and the surface-enhanced raman scattering spectrum signal of the subcutaneous tissue fluid includes inputting the permittivity, the physiological state and the surface-enhanced raman scattering spectrum signal of the subcutaneous tissue fluid as input features into a pre-trained multi-mode fusion calculation model, and inferring the insulin concentration in the subcutaneous tissue fluid of the user through the multi-mode fusion calculation model. In an embodiment of the first aspect of the disclosure, the method further comprises obtaining a near infrared spectrum signal measured by a sensing module attached to the skin surface of a user, determining the blood glucose concentration in subcutaneous tissue fluid of the user according to the near infrared spectrum signal, wherein the determining method of the blood glucose concentration comprises the steps that input features of the multi-mode fusion calculation model further comprise the near infrared spectrum signal, a result obtained by reasoning the multi-mode fusion calculation model further comprises the blood glucose concentration in subcutaneous tissue fluid of the user, the multi-mode fusion calculation model is a multi-task learning model, a variation trend of the ratio is obtained by fitting based on the ratio of the insulin concentration in subcutaneous tissue fluid to the blood glucose concentration at each moment, and the insulin resistance level of the user is determined according to the variation trend. In an embodiment of the first aspect of the disclosure, the multi-mode fusion calculation model further includes real-time compensating for a time difference between the tissue fluid and the insulin and/or blood glucose concentration in the blood according to the blood flow perfusion index, so as to obtain the insulin concentration and/or blood glucose concentration in the blood. In an embodiment of the first aspect of the disclosure, the first calibration is performed on the insulin concentration and/or the blood glucose concentration in the blood, wherein the first calibration method includes correcting the insulin concentration an