Advanced

Optimizing Diabetes Simulation Model Parameters using an Evolutionary Computing Approach

Li, Yanchen (2015) BINP31 20142
Degree Projects in Bioinformatics
Popular Abstract
Diabetes mellitus, especially type 2 diabetes, continuously becomes a global concern as this chronic metabolic disease which leasds over time to serious damage to heart, blood vessels, eyes, kidneys and nerves. It causes 1.5 million deaths each year according to a report by the World Health Organisation. It is a very complex disease that requires diverse treatment depending on patients and their health conditions. Personalized medication for patients with diabetes offers great potential to enhance the therapeutic effect by receiving personalized treatment regimens of already approved medications and other treatment options.
However, a central issue is the data availability to test and evaluate global optimization models covering all... (More)
Diabetes mellitus, especially type 2 diabetes, continuously becomes a global concern as this chronic metabolic disease which leasds over time to serious damage to heart, blood vessels, eyes, kidneys and nerves. It causes 1.5 million deaths each year according to a report by the World Health Organisation. It is a very complex disease that requires diverse treatment depending on patients and their health conditions. Personalized medication for patients with diabetes offers great potential to enhance the therapeutic effect by receiving personalized treatment regimens of already approved medications and other treatment options.
However, a central issue is the data availability to test and evaluate global optimization models covering all patients and all types of medications for a specific population. The sparseness of data, which is a consequence of practical limitations such as cost for lab tests, cost for acquisition of blood samples from patients, access to patients willing to contribute, etc., becomes the biggest problem.
In this thesis, a framework for disease simulation based on biological structure of human blood circulatory system is created and validated in order to increase the data resolution by efficiently interpolate data. A parallel genetic algorithm is introduced to fit patients’ clinical data with simulation and accelerate the calculation.
The results show a very high accuracy of prediction with the simulation for both glucose and insulin curve during the glucose test. The data resolution is increased to one point per minute which provide a great help for further optimization. This thesis also shows a potential for such framework to be modified to apply to extensive complexity of this personalization project.

Supervisor: Lars Hård
Master´s Degree Project 45 credits in Bioinformatics 2015
Lund University, Department of Biology (Less)
Please use this url to cite or link to this publication:
author
Li, Yanchen
supervisor
organization
course
BINP31 20142
year
type
H2 - Master's Degree (Two Years)
subject
language
English
id
8052292
date added to LUP
2015-10-06 16:16:03
date last changed
2015-10-06 16:16:03
@misc{8052292,
  author       = {Li, Yanchen},
  language     = {eng},
  note         = {Student Paper},
  title        = {Optimizing Diabetes Simulation Model Parameters using an Evolutionary Computing Approach},
  year         = {2015},
}