Bonfring International Journal of Industrial Engineering and Management Science
Online ISSN: 2277-5056 | Print ISSN: 2250-1096 | Frequency: 4 Issues/Year
Impact Factor: 0.541 | International Scientific Indexing(ISI) calculate based on International Citation Report(ICR)
Simple and Efficient Blood Glucose Measurement Technique Using Non Invasive Artificial Intelligence
Abstract:
At present, roughly about 50.8 million people in the world suffering from diabetes and maintaining the blood sugar level is extremely important. Diabetic patients need to frequently check their glucose level by means of pricking the finger for taking blood samples 4-5 times a day and maintain their sugar level through the process of taking appropriate dosage of medicine. In this paper, an artificial intelligence based image processing technique has been formulated, which non-invasively computes the glucose level exist in the urine sample of a person and therefore the corresponding blood glucose level of that person is computed. Blood sugar level of a person from his/her urine sample has been examined by observing the colour transformation of the test sample, when it is reacted with Benedict?s reagent. The colour transformation of the sample is recognized with the assistance of the camera and shows the result in the form of hue* value. This measurement can be effectively carried out by training the neural network with the hue value as the input vector and the glucose value as the test vector. Results show that the linear relationship has been obtained successfully with accuracy in the range of 96.93%.
Keywords: Benedicts Reagent, Colour Sensor, Neural Network, HSI Colour Format, Diabetes Mellitus.
Volume: 7 | Issue: 1
Pages: 09-13
Issue Date: March , 2017
DOI: 10.9756/BIJIEMS.8320
 |
|