Desarrollo e implementación de un prototipo de glucómetro no invasivo mediante espectrometría de infrarrojo cercano

Abstract

In the final degree work, a non-invasive measurement prototype was developed and implemented using the near infrared spectroscopy technique. Based on the theoretical study of near infrared spectroscopy, wavelengths, elements, and optical arrangements were determined for the development of hardware and software that allowed non-invasive glucose measurements on the right index finger. An invasive glucometer was used to relate the measurements of the implemented system with the blood glucose values. In this way, a database that contained information on the glucose concentration based on the absorbance of the measured signals was prepared. The database was used in the training of a set of neural networks, which were implemented in system programming to perform validation tests. In the validation tests, 54 measurements were made with the implemented system and they were compared with the measurements made by an invasive glucometer. These results were evaluated using evaluation indices used in glucose measurement systems. Additionally, the calibration results of the system and the effects on the measurements produced by the factors related to the near infrared spectroscopy technique were presented.