Attention deficit hyperactivity disorder (ADHD) is a childhood behavioral disorder that can persist into adulthood. Electroencephalography (EEG) plays a significant role in assessing the neurophysiology of ADHD because of its ability to reveal complex brain activity. The present study proposes an EEG-based diagnosis system using the phase space reconstruction technique to classify ADHD and control adults. Electric activity is recorded for 47 ADHD and 50 control adults during the eyes-open, eyes-closed, and Continuous Performance Test (CPT) condition. Various statistical features are extracted from Euclidean distances based on phase space reconstruction of signals. The proposed system is evaluated with 2 feature selection methods (correlation-based feature selection and particle swarm optimization) and 5 machine learning methods (neural dynamic classifier, support vector machine, enhanced probabilistic neural network, k-nearest neighbor, and naive-Bayes classifier). Experimental results showed the highest testing accuracy of 93.3% under the eyes-open, 90% under the eyes-closed, and 100% under the CPT condition. This study focused on the utility of phase space reconstruction of brain signals to discriminate between ADHD and control adults.
An Ultrashort-Term Net Load Forecasting Model Based on Phase Space Reconstruction and Deep Neural Network