Please use this identifier to cite or link to this item:
Title: Dynamic modeling of type 1 diabetic metabolism
Authors: Jivesh Ramduny
Keywords: DRNTU::Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence
Issue Date: 2015
Abstract: Diabetes Mellitus is a metabolic disorder whose prevalence is as high as 387 million people globally. This number accounts for 5.5% of the world population and has outpaced the estimated diabetic subjects by 2030 by 11 million subjects. Today, the occurrences of Type 1 Diabetes Mellitus (T1DM) encompass 5-10% of all cases of Diabetes Mellitus. Since Type 1 Diabetes Mellitus (T1DM) happens in 1 out of every 600 children, its etiology is paramount. The research presented in the Final Year Project (FYP) report has attempted to mimic the metabolism of a Type 1 Diabetes Mellitus (T1DM) subject using real human clinical data from the Center of Endocrinology of the KK Women’s and Children’s Hospital. In the process, the focus was shifted to address the present insulin therapy of manual insulin infusion. The methods approached in the course of the research embraced the extensive understanding of the operations of Genetic Algorithm. Genetic Algorithm enabled the parameters to self-force their values in order to reach convergence and resulted in the optimal value of the glucose utilization rate. Further experimental benchmarks were conducted which capitalized on the fitting of the predicted blood glucose levels given an amount of insulin bolus after a meal declaration. Careful analysis was performed with the incorporation of a new fitness function which dealt with the merging of correlation and the mean squared error of the predicted and actual concentration of blood glucose levels at any recorded time. At the end of the experimental phase, an extensive comparison was conducted to illustrate the performances of the Modified GlucoSim which is an existing computational intelligence model in accordance of W.R.Puckett (1992) thesis and the applied Genetic Algorithm relative to the clinical data of the subject. Although the research dealt with one particular set of clinical data of a diabetic subject, the experimental phase provided substantial information which laid the foundation for future research on insulin infusion. However, a sound understanding of the dynamics of Type 1 Diabetes Mellitus (T1DM) is needed to improve the techniques applied.
Rights: Nanyang Technological University
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:SCSE Student Reports (FYP/IA/PA/PI)

Files in This Item:
File Description SizeFormat 
  Restricted Access
3.29 MBAdobe PDFView/Open

Google ScholarTM


Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.