ABSTRACT
Advances in information technology and data collection methods have enabled high-speed collection and storage of huge amounts of data. Data mining can be used to derive laws fromlargedata volumes and their characteristics [1]. Similarly, fuzzy logic by facilitating the understanding of events is considered a suitable complement to scientific data mining. The present study used clustering to identify the independent characteristics of data. Related fuzzy sets, linguistic variables, and data classifications were defined, and the index was introduced based on the characteristics extracted from useful results. By considering the disease risk factors, the results were analyzed. Two factors contributing to the health improvement or deterioration were defined: 'age' and 'the appropriateness or inappropriateness between insulin level and blood sugar'. In addition, according to the results, fasting had a positive effect on fatty substances of the blood [cholesterol and triglycerides]. The results can help us determine whether or not an individual with a cardiovascular disease should fast in the month of Ramadan. However, due to variations in some features such as blood pressure throughout the day, there are uncertainties in some input data; therefore, the results could be far from reality. If it is possible to generate fuzzy data, then we can obtain more accurate results
ABSTRACT
Epilepsy is a clinical syndrome in which seizures have a tendency to recur. Sodium valproate is the most effective drug in the treatment of all types of generalized seizures. Finding the optimal dosage [the lowest effective dose] of sodium valproate is a real challenge to all neurologists. In this study, a new approach based on Adaptive Neuro-Fuzzy Inference System [ANFIS] was presented for estimating the optimal dosage of sodium valproate in IGE [Idiopathic Generalized Epilepsy] patients. 40 patients with Idiopathic Generalized Epilepsy, who were referred to the neurology department of Mashhad University of Medical Sciences between the years 2006-2011, were included in this study. The function Adaptive Neuro- Fuzzy Inference System [ANFIS] constructs a Fuzzy Inference System [FIS] whose membership function parameters are tuned [adjusted] using either a backpropagation algorithm alone, or in combination with the least squares type of method [hybrid algorithm]. In this study, we used hybrid method for adjusting the parameters. The R-square of the proposed system was%598 and the Pearson correlation coefficient was significant [P <0.05] and equal to 0.77, but theT-test was not significant [P >0.05]. Although the accuracy of the model was not high, it wasgood enough to be applied for treating the IGE patients with sodium valproate. This paper presented a new application of ANFIS for estimating the optimal dosage of sodium valproate in IGE patients. Fuzzy set theory plays an important role in dealing with uncertainty when making decisions in medical applications. Collectively, it seems that ANFIS has a high capacity to be applied in medical sciences, especially neurology