Introduction: A full investigation into the features extracted from voice signals of people with and without Parkinson's disease was performed. A total of 31 people with and without the disease participated in the data collection phase. In this study subjects' voice signals were used to let computer decide whether the person is suffering from the disease or they are not. Methods: Their voice signals were recorded and processed. The relevant features were then extracted. Features were fed to different classifiers so as to be let them decide whether the subjects have the disease or not. Three different classifiers were used in order to rule out any doubt about the validity of classification performance on the given data. Results: The use of a variety of feature selection methods resulted in a good performance for the diagnosis of Parkinson. The classifiers' performances were compared with one another and showed that the best performance was obtained with a correct rate of 0.9382 when using the KNN classifier. Discussion: Results reveal that the use of proposed feature selection method results in a desirable precision for the diagnosis of Parkinson's disease (PD). The performances were assessed from different points of view, providing different aspects of the diagnosis, from which the physicians are able to choose one with higher accuracy in the diagnosis.
|Number of pages||9|
|Journal||Basic and Clinical Neuroscience|
|Publication status||Published - 1 Dec 2011|
Bibliographical noteCopyright the Author(s) 2011. Version archived for private and non-commercial use with the permission of the author/s and according to publisher conditions. For further rights please contact the publisher.
- feature selection
- Parkinson's Disease (PD)