Insulin chart prediction for diabetic patients using hidden markov model (HMM) and simulated annealing method

Abstract

Most of the diabetic patients need to take insulin before every meal. The doctors have to decide insulin doses for every patient according to the patient’s previous records of doses and sugar levels measured at regular intervals. This paper proposes a hidden Markov model to predict the insulin chart for a patient and uses simulated annealing search algorithm to efficiently implement the model. The onemonth chart maintained by the patient has been used to train the model, and the prediction for next fifteen days is done on the basis of the trained data.We discussed the results with the university medical doctor; he was very pleased to see to the result obtained.