Fig. 1.

Schematic of the model performance validation procedure. During leave-one-out cross-validation (LOOCV), we first isolated the network of difference between IGD_ADHD+and IGD_ADHD−groups using NBS to elucidate a network related to comorbid ADHD. Next, FA values of each network edge were extracted and defined as predictors of the regression model. The PLS regression model was trained on data and tested with one left-out participant. The test was performed iteratively for all subjects (YIAS, Young Internet Addiction Scale; K-ARS-P, Dupaul's ADHD scale-Korean version).