Loading...

Design & Identification of a Fuzzy Model to Predict Dynamic Trends of Blood Glucose Level in Type I Diabetic Patients

Salehi, Samane | 2014

535 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 46187 (06)
  4. University: Sharif University of Technology
  5. Department: Chemical and Petroleum Engineering
  6. Advisor(s): Bozorgmehri, Ramin
  7. Abstract:
  8. Nowadays diabetes is considered as one of the common diseases among mankind. Different body organs utilize the glucose or the sugar existed in various kinds of food as an energy source in order to provide required energy for their activities. Insulin hormone which is created by the pancreas gland has a very significant role in easing the respective procedure. Therefore, when the body is not capable of producing enough insulin, then glucose aggregation in the blood leads to higher blood sugar and thus a disease called diabetes type1. Up to now, considerable research activities have been done regarding the diabetes type1modelling. One of the bolded drawbacks of such models is that the effect of some parameters involved in blood sugar rate variation is ignored and they are not calculateable in the field of mathematical problems. Based on this, in the current research, the use of fuzzy logic approach to employ such parameters’ effects is addressed. Examples of such parameters are: physical activity, weight, gender, air pollution, diseases and neural pressure or stress. As a result, based on the mentioned parameters, the patient’s blood sugar rate in 2 hours is predicted. This resulted value besides the injected insulin, received pure sugar, patient’s sugar and insulin amounts during the previous hours predict the patient’s blood sugar dynamic variations based on fuzzy and recursive fuzzy systems regulations. At the end, the results are compared with Cinar physiological model results. The experiment on the model is done with various amounts of insulin with the emphasis on its various amounts and also injection time before, at the same time and after the injection. Results confirm the accuracy of the proposed model regarding their accepted coincidence. On the other side, to evaluate the model behavior for the effects of other parameters, two experiments are done by applying a lot of, medium and a few activities in two conditions of injecting the insulin in accordance to the used carbohydrate and not insulin injecting and the results are in consistence with medical predictions
  9. Keywords:
  10. Diabetes Mellitus Type 1 ; Fuzzy Logic ; Diabetes Modelling ; Recursive Fuzzy Systems

 Digital Object List

 Bookmark