Maryam Sadat Fadavi - Faculty of Electrical Engineering, K.N. Toosi University of Technology, Tehran, Iran
Fatemeh Hasanzadeh - Faculty of Electrical Engineering, K.N. Toosi University of Technology, Tehran, Iran
Maryam Mohebbi - Faculty of Electrical Engineering, K.N. Toosi University of Technology, Tehran, Iran
Peyman Hasani Abharian - Institute of Cognitive Science Studies, Tehran, Iran

Opioid addiction is one of the serious global problems. Currently, the detection of opioid addiction is performed by tests that can have misdiagnosis and tested cases can change the results of these tests by some tricks. To develop a more impervious method, EEG signals can be helpful. To detect opioid addicts, the 19-channel EEG signal of 22 addicted patients and 22 control subjects are recorded. The control group includes drug abstinence individuals. The recurrence plot (RP) analysis is performed on EEG signals in the delta, theta, alpha1, alpha2, beta1 and beta2 frequency bands. The SVM classifier applied to features obtained from RP analysis to discriminate opioid addicts from control subjects. The combination of features from all studied frequency bands led to 93.02% accuracy, 95.5% specificity and 90.9% sensitivity. The obtained high accuracy indicates that promising results in addiction detection can be obtained the proposed method.

Addiction, EEG signal, Drug abuse, Recurrence plot analysis, Classification