ROMJIST Volume 25, No. 1, 2022, pp. 100-113
Mustafa GERGER, Abdülkadir GÜMÜŞÇÜ Diagnosis of Parkinson's Disease Using Spiral Test Based on Pattern Recognition
ABSTRACT: Recently, Parkinson's disease has become one of the most serious health problems. The incidence is significantly increased, especially in people over 65 years of age. The adverse effects of the disease on motor activities distract patients from social environments. Early diagnosis is important in Parkinson's disease. In addition, the diagnosis of Parkinson's disease is quite difficult. Therefore, computer-aided systems are used. The Parkinson’s spiral drawing test is one of these methods. Many researchers have classified these drawings with different approaches. In this study, the spirals drawn in the diagnosis of Parkinson's disease were analyzed by pattern recognition methods. Patient and control group drawings were assumed to be characters. Thus, 123.066 features were obtained for each drawing. In this way, a high p low n problem occurs when the data set is converted. In order to overcome this problem, feature selection process with genetic algorithm was applied as a pre-process. k-nearest neighbor and decision trees were used for the classification process. The results are validated by a leave one out cross validation method. In this study, the data set prepared by the Department of Neurology in Cerrahpasa Faculty of Medicine, Istanbul University was used. The experimental results obtained show that the proposed method gives successful results. Specifically, accuracy value obtained 1.00 in the decision tree classification by selecting the feature with the genetic algorithm. These results showed that the proposed method can be used in the diagnosis of Parkinson's disease.KEYWORDS: Artificial Intelligence; Medical Informatics; Parkinson’s diagnosis; Spiral test; Pattern recognition; Classification; Genetic algorithmRead full text (pdf)