Non Invasive Platform to Estimate Fasting Blood Glucose using Salivary Electrochemical Parameters

Sarul Malik1, Shalini Gupta2, Sneh Anand2

  • 1PhD student Indian Institute of Technology
  • 2Indian Institute of Technology Delhi

Details

Category

1-Page Extended Abstract (Poster)

Sessions

13:25 - 14:15 | Thu 16 Feb | Ballroom D | ThRPF

Rapid Fire Session 02: Sensor Informatics I

Abstract

Elevated blood glucose is the prime indicator of diabetes. We have developed a non invasive platform to estimate fasting blood glucose level (FBGL) using salivary electrochemical parameters (SEP). A proof of concept study has been performed with the sensor setup interfaced with various machine learning algorithms (MLA). Neural net boosting regression estimated FBGL with an accuracy of 87.6%, mean relative deviation of 15.0% and coefficient of determination as 0.76.

Additional Information

No information added

Video

No videos found