EEG Source Imaging Based on Dynamic Sparse Coding as ADHD Biomarker

The study of the psychiatric disorder denominated Attention-deficit hyperactivity disorder (ADHD) demands the assessment of specific behavior, measured and evaluated through biomarkers like the neuroimaging that is applied due to the assumed association between with changes in the structure and function of the ADHD brain. Because of the provided time resolution, Electroencephalographic (EEG) signals and derived versions have recently gained increased attention for studying event-related potentials (ERPs). Moreover, relate to the ADHD diagnosis, techniques of EEG/ERP source imaging (ESI) are effective to locate brain areas related to attention task and analyze spatiotemporal patterns of the P300 wave. Therefore, with the aim to accurately determine the spatial location and temporal patterns involved in attention task, there is a need for implementing an adequate ERP marker able to incorporate the spatial and temporal prior information to the ESI solution. In this paper, the influence of the source reconstruction is evaluated on visual and auditory evoked potentials through an ESI solution, namely, Dynamic Sparse Coding that is based on physiological motivated spatio-temporal constraints over the source representation. As a result, the DSC-based approach improves the characterization of the spatio-temporal dynamics of the attentional evoked potentials processes, including reduced amplitudes in the P300 components of the ERPs in the ADHD group.