Volume 3, Issue 2 (Spring 2017)                   Caspian.J.Neurol.Sci 2017, 3(2): 106-117 | Back to browse issues page

XML Print

Department of Clinical Psychology, Islamic Azad University, Science and Research Branch, Tehran, Iran; babak2000_m@yahoo.com
Abstract:   (1119 Views)
Background: Identifying mental disorder biomarkers is one of the leading goals of the clinical science.
Objectives: This study aimed to provide an artificial intelligence based solution and software program to diagnose the type and severity of mental disorders according to the quantitative electroencephalogram (QEEG) of patients. Materials and Methods: The QEEG data collected from 45 patients addicted to one of the substances (crystal-glass methamphetamine [n=15], tramadol [n=15], heroin/opium [n=15]) and 15 healthy people. They were entered into SPSS version 20 and analyzed by Discriminant Analysis (DA) function and simultaneously used as the Training Group of the artificial neural network (ANN) of the diagnosis software. In order to test and validate the software, in the following, QEEG was also recorded from the remaining 60 subjects (45 addicted and 15 healthy people).
Results: The results obtained from the software were 0.836, 0.884, 7.21, 0.19, 0.712, and 0.890, respectively. Meanwhile, the values of these parameters for DA were 0.677, 0.66, 1.99, 0.49, 0.363, and 0.739, respectively. The results of the software significantly improved the diagnosis. Totally nine discriminant functions were obtained for the frontal, parietal and central lobes was obtained according to the delta, Theta, Alpha and Beta variables.
Conclusion: As a result, intelligent diagnosis software provided can be used with a high sensitivity and great specificity rather than Paper-Pencil tests for accurate diagnosis of the type of disorder and expressing its severity at a confidence level that is scientifically computed and displayed.
Full-Text [PDF 1182 kb]   (454 Downloads) |   |   Full-Text (HTML)  (239 Views)  
Type of Study: Research | Subject: Special
Received: 2017/08/7 | Accepted: 2017/08/7 | Published: 2017/08/7