Innovative measurement of blood glycemic level from saliva for diabetic patients using thermal image technique and analyzing through fractional wavelet transform compared with invasive measurement of blood glucose using glucometer.

In this paper, the Blood Glycemic level (BG) is measured using saliva as a sample through Fractional Wavelet Transform (FRWT) based on thermal properties and compared with an invasive method. Thermal image of saliva is captured with a thermal camera for 10 subjects(5-Diabetes,5-Non Diabetes) and BG is measured through FWT in image processing. In this proposed work, it is observed that non-invasive linear regression has a higher accuracy of 90% than invasive linear regression, which has an accuracy of 86.8 percent and significance (<0.01) which has been calculated using the SPSS tool. Within the limitations of this study, BG measurement using saliva through FRWT has a higher significance than an invasive technique. [ABSTRACT FROM AUTHOR]