Measurement and classification of urinary dielectric properties for type 2 diabetes mellitus and chronic kidney disease / Mun Peck Shen

Mun, Peck Shen (2016) Measurement and classification of urinary dielectric properties for type 2 diabetes mellitus and chronic kidney disease / Mun Peck Shen. PhD thesis, University of Malaya.

Preview

PDF (Thesis PhD) Download (2514Kb) | Preview

Abstract

Diagnosis and monitoring of diabetes and chronic kidney disease are of crucial importance for preventing end-stage kidney failure. The measurement of dielectric properties has generated interest for clinical utility. In this study, the urinary dielectric properties and behaviour of subjects with Type 2 diabetes mellitus (DM), subjects with chronic kidney disease (CKD), and normal subjects are investigated. The measurements were conducted using open-ended coaxial probe at microwave frequencies between 0.2 GHz and 50 GHz at room temperature (25°C), 30°C and human body temperature (37°C), respectively. The measurement of urinary dielectric properties for DM subjects that obtained dielectric constant increased with glycosuria level of more than 5 g/L at low frequencies and correlated positively with glycosuria level at frequencies above 40 GHz. Loss factor correlated negatively with glycosuria level at frequencies above 15 GHz. The strongest statistically significant difference in urinary dielectric properties was reported at room temperature (25°C) and body temperature (37°C) across different glycosuria and proteinuria levels, respectively. Statistically significant differences were found in the urinary dielectric properties of the CKD subjects compared to those of the normal subjects. Urinary dielectric properties correlated positively and negatively with proteinuria level at frequencies below and above the ―cross-over‖ frequency point, respectively. The experimental data closely matched the single-pole Debye model. The relaxation dispersion and relaxation time increased with the glycosuria and proteinuria level, while decreased with the temperature. Classifications of urinary dielectric properties were conducted using support vector machine (SVM). In two-group classifications, the highest accuracy of 88.72% was obtained by differentiating DM subjects from normal subjects. The highest accuracy was achieved at 67.62% for three-group classifications. The best classification accuracies were obtained at 30°C. This study demonstrated the potential diagnostic and prognostic value of urinary dielectric properties for Type 2 DM and CKD.

Actions (For repository staff only : Login required)