An auditory brainstem response-based expert system for ADHD diagnosis using recurrence qualification analysis and wavelet support vector machine

Attention Deficit Hyperactivity Disorder (ADHD) is a common disorder in children. Due to lack of suitable biomarker or test, diagnosis of ADHD children is complicated and needs comprehensive evaluations. Evidences show that, ADHD children have deficit in their brainstem timing and cortex auditory processing. We assessed their auditory brainstem response to speech stimuli. Due to nonlinear and dynamic characteristics of biological signals they should be represented by features that are based on their nature. In this study wavelet coefficients and recurrence qualification analysis features were used to represent signals in a comprehensive way. In this article, we addressed the problem of discrimination of ADHD children from Normal. Wavelet Support Vector machine with Mexican hat and Morlet kernels were used in order to classifying these children. Our method demonstrated %98.57 classification accuracy.