Blood Glucose Measurement Based on Infra-Red Spectroscopy

TL;DR

This paper presents a new algorithm based on PLS regression for measuring blood glucose levels using infrared spectroscopy, achieving accuracy comparable to standard invasive glucometers.

Contribution

The study introduces an optimized PLS regression algorithm for non-invasive blood glucose measurement via infrared spectroscopy.

Findings

Mean estimation error of 7 mg/dL

Results agree well with standard glucometers

Infrared spectroscopy can reliably measure glucose levels

Abstract

An algorithm based on PLS regression has been developed and optimized for measuring blood glucose level using the infra-red transmission spectrum of blood samples. A set of blood samples were tagged with their glucose concentration using an accurate invasive glucometer and analyzed using a standard FTIR spectrometer. Using the developed algorithm, the results of the FTIR spectroscopy of the samples were analyzed to find the glucose concentration in the samples. The obtained glucose concentration by the algorithm were in good agreement with the results obtained by the standard glucometer, and the mean estimation error was 7 mg/dL. This error is in the range of available commercial invasive meters.