Non-Invasive antenna sensor based continuous glucose monitoring using pancreas dielectric radiation signal energy levels and machine learning algorithms

In this paper, H shape with I slot patch (HSIS) antenna is utilized as sensor for proposed continuous glucose monitoring (CGM) device. HSIS antenna sensor is placed on pancreas of human body and acquires dielectric radiation signals from the pancreas. Pancreas dielectric property changes during insulin secretion. HSIS antenna as a passive sensor acquires pancreas dielectric radiation signals from pancreas more effectively due to H shape and I slot. H shape antenna covers maximum region on human pancreas. During signal acquisition, crumbling effect causes bending and stretching in pancreas and leads to noise in signal, which is prevented through H-shape in HSIS antenna. Traditional CGM method measures diabetic value through invasive sensors. Non-invasive diabetic measurement device never measures CGM because of sensor mount for long duration on human body which is the major problem and leads to infections due to pricked sensor on the body for more than 14 days. In this paper, pancreas is identified as optimized location for CGM through HSIS antenna, among other human body parts such as stomach, carpus and thumb finger. Optimized location is identified through return loss and VSWR values. HSIS antenna acquired pancreas dielectric radiation signals process with Rational Dilation Wavelet Transforms (RADWT). Energy band and statistical values are obtained from signals and correlated with labo-ratory diabetic values. Experimental study of CGM is performed in 150 diabetic persons. The proposed HSIS antenna as sensor shows the fluctuation in diabetic value through signals and predict the diabetic value with 91.85% accuracy, when compared to laboratory based diabetic value measurement.