Method for non-invasive glucose measurement

Abstract

A non-invasive blood glucose measurement method includes sequentially illuminating a test site with a light source and sensing the transmitted or reflected light from the test site using a photosensitive device to generate a photoelectric data sequence. The method further includes converting the photoelectric data sequence into a photovolume change map using a processing device; performing a first phase synchronization processing on the photovolume change map to obtain a phase-synchronized photovolume change map and phase information; performing a second phase synchronization processing on the photoelectric data sequence based on the phase information to obtain a phase-synchronized photoelectric data sequence; merging the phase-synchronized photoelectric data sequences into a phase-synchronized multispectral photoelectric data sequence; selecting photoelectric data segments from the phase-synchronized multispectral photoelectric data sequence; extracting photoelectric data feature representations from the photoelectric data segments; and estimating blood glucose concentration based on the photoelectric data feature representations.

Claims

1. A non-invasive blood glucose measurement method, comprising: sequentially illuminating a test site of a subject with multiple light sources of different wavelengths, wherein each light source is continuously activated for a predetermined time; sensing light transmitted or reflected through the test site using a photosensitive device to generate multiple photoelectric data sequences corresponding to the different wavelengths; and performing the following steps using a processing device: converting each photoelectric data sequence into a photovolume change profile; performing a first phase synchronization process on the photovolume change profile to obtain multiple phase-synchronized photovolume change profiles and phase information corresponding to the phase-synchronized photovolume change profiles; performing a second phase synchronization process on the photoelectric data sequences based on the phase information to obtain multiple phase-synchronized photoelectric data sequences; merging the phase-synchronized photoelectric data sequences into a phase-synchronized multispectral photoelectric data sequence; selecting a photoelectric data segment from the phase-synchronized multispectral photoelectric data sequence; extracting a photoelectric data feature representation from the photoelectric data segment; and estimating a blood glucose concentration of the subject based on the photoelectric data feature representation.

2. The blood glucose measurement method as described in claim 1 further comprises: extracting multiple time-domain features from the aforementioned phase-synchronous photovolume change chromatogram; and estimating the blood glucose concentration of the subject based on the aforementioned photoelectric data feature representation and the aforementioned time-domain features.

3. The blood glucose measurement method as described in claim 1, wherein converting each of the aforementioned photoelectric data sequences into the aforementioned photovolume change mapping further comprises: selecting a region of interest (ROI) from the aforementioned photoelectric data sequence; calculating a time-series signal formed by the change of signal intensity in the aforementioned ROI over time; and performing a filtering process on the aforementioned time-series signal to generate the aforementioned photovolume change mapping.

4. The blood glucose measurement method as described in claim 1, wherein the different wavelengths are between 800 nm and 1000 nm.

5. The blood glucose measurement method as described in claim 1 further comprises: receiving multiple reference blood glucose concentrations obtained by the subject from multiple previous measurements using a reference blood glucose measuring device; and correcting a regression model used to estimate the blood glucose concentrations using multiple historical photoelectric data features corresponding to the multiple previous measurements and the reference blood glucose concentrations.

6. A non-invasive blood glucose measurement method, comprising: generating mixed polarized light through multiple light sources with different wavelengths to simultaneously irradiate a test site of a subject; setting a first polarizer to allow the mixed polarized light to pass through the first polarizer to generate polarized light, and allowing the polarized light to be incident on the test site; setting a second polarizer to allow reflected light from the test site to pass through the second polarizer to filter out reflected light from the skin surface and obtain filtered reflected light; sensing the filtered reflected light through a multispectral light sensing device to generate a phase-synchronized multispectral photoelectric data sequence; and performing the following steps through a processing device: selecting a photoelectric data segment from the phase-synchronized multispectral photoelectric data sequence; extracting a photoelectric data feature representation from the photoelectric data segment; and estimating a blood glucose concentration of the subject based on the photoelectric data feature representation.

7. The blood glucose measurement method as described in claim 6 further comprises: converting the aforementioned phase-synchronized multispectral photoelectric data sequence into a multispectral photovolume change profile; extracting multiple time-domain features from the aforementioned multispectral photovolume change profile; and estimating the aforementioned blood glucose concentration of the subject based on the aforementioned photoelectric data feature representation and the aforementioned time-domain features.

8. The blood glucose measurement method as described in claim 7, wherein converting the aforementioned phase-synchronized multispectral photoelectric data sequence into the aforementioned multispectral photovolume change mapping further comprises: selecting a region of interest corresponding to each of the aforementioned different wavelengths of photoelectric data from the aforementioned phase-synchronized multispectral photoelectric data sequence; calculating a time-series signal formed by the change of signal intensity in the aforementioned region of interest over time; and performing filtering processing on the aforementioned time-series signal to generate the aforementioned multispectral photovolume change mapping.

9. The blood glucose measurement method as described in claim 6, wherein the polarization directions of the first polarizer and the second polarizer are perpendicular to each other.

10. The blood glucose measurement method as described in claim 6 further comprises: receiving multiple reference blood glucose concentrations obtained by the subject from multiple previous measurements using a reference blood glucose measuring device; and correcting a regression model used to estimate the blood glucose concentrations using multiple historical photoelectric data features corresponding to the multiple previous measurements and the reference blood glucose concentrations.

Images