Evaluation of photic effect-due electroensefalographic changes in internet addiction in adolescents with machine learning

Introduction and Objective: In this study, it was aimed to evaluate the neurodevelopmental Electroencephalographic changes of the 10-19 age group patients who applied to the Dicle University Faculty of Medicine, Pediatrics and Adolescent Outpatient Clinic, and their addiction levels, using Machine Learning. It is thought that this study may help to illuminate the pathophysiology and treatment of internet addiction. Material and Method: 30 adolescent patients with internet addiction and 6 patients without internet addiction were included in the study. EEG imaging was performed on these patients, and the determination of effective EEG Channels with Artificial Neural Networks and the evaluation of data with Machine Learning were performed. In addition, categorical variables were compared between groups using the chi-square test. Results: Internet addiction was found more frequently in males, between the ages of 10-13, with normal body weight and height Z scores, in those with a low income level, and in those who use tablets. Among the IA cases, mobile phone use in the 14-18 age group was found to be significantly higher in the 10-13 age group. Using artificial neural networks and classification performance, it was determined that brain activity reflections differ in adolescence according to electrode locations. As visual findings related to sick and healthy individuals in adolescence, for the delta band, the brain dynamics of the control group have a more complex characteristic, EEG complexity in dependent adolescents differed significantly from the control group as it went towards higher bands, Brain activity of individuals with IA is significantly more complex in the Gamma and Beta bands, in the alpha band, differences between the two groups were found to be minimal. It was determined that while the brain activities of healthy individuals were more complex in the low frequency region, the brain activities of IA individuals exhibited more complex dynamics in the high frequency region. Conclusion: Noninvasive methods (EEG etc.) will play an important role in investigating the neurobiological mechanism and in the treatment of IA. Multiple imaging techniques with behavioral measures should be necessary to improve our understanding of IA. This may aid in understanding the neural mechanisms underlying inhibitory control deficits in IA and pointing out possible future interventions.