Design of Breath Acetone Detection for Diabetes Detection and Classification using Electronic Nose

This study aims to develop an affordable, user-friendly, and non-invasive system for diabetes detection and classification and measuring blood glucose levels using breath analysis. The system consists of a combination of Raspberry Pi and Arduino Uno, to develop a portable Android-based system. The research explores the correlation between breath acetone levels and diabetes, using the electronic nose (E-Nose) technology. It implements a Convolutional Neural Network (CNN) and Support Vector Machine (SVM) to establish connections between measured variables and blood sugar levels. The results of the study will provide insights regarding the system's capability to precisely categorize and identify diabetes, as well as the effectiveness of various machine learning algorithms employed for this objective. These findings will be compared with the reference standard test for a diabetes diagnosis to ascertain the system's accuracy. If the results are promising, the study will provide proof of concept for a non-invasive., low-cost, and easy-to-use diabetes screening tool that can be used in resource-limited settings.