Biomarker value calculation method
A computer-implemented method to improve the accuracy of a calculation of a biomarker value from a spectral measurement. The computer-implemented method comprises receiving a primary spectral measurement from a primary detector, receiving, a secondary measurement from a secondary detector, and calculating a value of the biomarker using a biomarker algorithm. The biomarker algorithm takes an input from the primary spectral measurement and a calibration parameter from the secondary measurement.
1. A computer-implemented method to improve the accuracy of a calculation of a biomarker value from a spectral measurement, the computer-implemented method comprising:
receiving a primary spectral measurement from a primary detector,
receiving a secondary measurement from a secondary detector, and
calculating a value of the biomarker using a biomarker algorithm, which takes an input from the primary spectral measurement and a calibration parameter from the secondary measurement,
wherein the secondary measurement is a secondary spectral measurement that is less variable, with the biomarker, than the primary spectral measurement, and
wherein the method further comprises:
generating the biomarker algorithm using a plurality of training primary spectral measurements, a respective plurality of training secondary measurements, and a respective plurality of reference biomarker value labels; or
applying a calibration model to the secondary measurement, the calibration model taking an input from the secondary measurement, and the calibration model outputting the calibration parameter.
2. The computer-implemented method of claim 1 , wherein the secondary measurement is a spectral measurement.
3. The computer-implemented method of claim 2 , wherein the secondary measurement is in a different wavelength band to the primary spectral measurement.
4. The computer-implemented method of claim 2 , wherein the secondary measurement is a measurement of short wavelength infrared spectral data.
5. The computer-implemented method of claim 1 , wherein the calibration parameter is the secondary measurement.
6. The computer-implemented method of claim 1 , wherein the primary spectral measurement includes a measurement of red wavelength spectral data and a measurement of infrared wavelength spectral data.
7. The computer-implemented method of claim 6 , wherein the biomarker is oxygen saturation.
8. The computer-implemented method of claim 1 , wherein the primary spectral measurement is a measurement of short wavelength infrared spectral data.
9. The computer-implemented method of claim 8 , wherein the biomarker is a metabolite.
10. The computer-implemented method of claim 1 , wherein the biomarker algorithm is a trained machine learning model.
11. The computer-implemented method according to claim 1 , wherein the biomarker algorithm is a classification model, or a regression model, or a combination of a classification model and a regression model.
12. A device comprising a processor, the processor configured to carry out the method of claim 1 .
13. The computer-implemented method of claim 1 , wherein the secondary measurement is approximately constant with respect to the biomarker value, and is variable with respect to a user specific characteristic.
14. A computer-implemented method to improve the accuracy of a calculation of a biomarker value from a spectral measurement, the computer-implemented method comprising:
receiving a primary spectral measurement from a primary detector,
receiving a secondary measurement from a secondary detector, and
calculating a value of the biomarker using a biomarker algorithm, which takes an input from the primary spectral measurement and a calibration parameter from the secondary measurement,
wherein the method further comprises:
generating the biomarker algorithm using a plurality of training primary spectral measurements, a respective plurality of training secondary measurements, and a respective plurality of reference biomarker value labels; or
applying a calibration model to the secondary measurement, the calibration model taking an input from the secondary measurement, and the calibration model outputting the calibration parameter.
15. The computer-implemented method of claim 14 , comprising:
applying the calibration model to the secondary measurement, the calibration model taking the input from the secondary measurement, and the calibration model outputting the calibration parameter.
16. The computer-implemented method of claim 15 , wherein the calibration model is generated using a plurality of training secondary measurements and a respective plurality of calibration parameter labels,
wherein the plurality of training calibration parameters are calculated using a respective plurality of training primary spectral measurements and a respective plurality of reference values of the biomarker.
17. The computer-implemented method of claim 15 , wherein the calibration model is a trained machine learning model.
18. The computer-implemented method of claim 15 , wherein the calibration model is a classification model, or a regression model, or a combination of a classification model and a regression model.
19. The computer-implemented method according to claim 14 , further comprising:
generating the biomarker algorithm using the plurality of training primary spectral measurements, the respective plurality of training secondary measurements, and the respective plurality of reference biomarker value labels.
20. A computer-implemented method to improve the accuracy of a calculation of a biomarker value from a spectral measurement, the computer-implemented method comprising:
receiving a primary spectral measurement from a primary detector,
receiving a secondary measurement from a secondary detector, and
calculating a value of the biomarker using a biomarker algorithm, which takes an input from the primary spectral measurement and a calibration parameter from the secondary measurement,
wherein the biomarker is temperature, and
wherein the method further comprises:
generating the biomarker algorithm using a plurality of training primary spectral measurements, a respective plurality of training secondary measurements, and a respective plurality of reference biomarker value labels; or
applying a calibration model to the secondary measurement, the calibration model taking an input from the secondary measurement, and the calibration model outputting the calibration parameter.
21. The computer-implemented method of claim 20 , wherein the primary spectral measurement is a measurement of short wavelength infrared spectral data.