The Fronto-occipital Network in Pre-attentive Detection of Visual Changes: An Event-related Optical Signal (EROS) Study

Pre-attentive change detection refers to the ability of human brains to monitor and detect changes in the environment even without attention. Previous studies showed that the Inferior Frontal Cortex (IFC) and the Superior Temporal Cortex (STC) were part of the brain network in pre-attentive detection of auditory changes. Specifically, an early IFC followed by STC then late IFC (i.e., early IFC-STC-late IFC) activation pattern was observed in detecting ambiguous auditory changes. As proposed by the prediction violation framework, the early IFC is related to reinstating the prediction model before the change detection process in the STC; when a prediction violation is detected, STC generates a prediction error that triggers the prediction model updating process which is shown as the late IFC activity. However, it remains unclear whether this fronto-sensory network organization is specific to detecting auditory changes or can also be applied to detecting visual changes. To address this question, the current study investigates the fronto-occipital (OC) network in pre-attentive detection of visual changes by using the Event-related optical signal (EROS). As EROS measures the change in optical properties of the brain associated with neuronal activity, it is able to localize brain activities temporally and spatially in the millisecond and sub-centimeter range. Twenty-four participants were recruited and presented with a visual passive oddball paradigm consisted of a simple physical feature change (i.e., orientation change of a bar array) and an abstract regularity violation (i.e., violating the clockwise rotation pattern among the standard bar arrays). Based on the number of preceding standards, each of the physical and abstract change deviants were further divided into the short train and the long train conditions. A shorter standard train or an abstract regularity among the standard is less likely to provide sufficient information to establish a stable prediction model. In detecting physical changes, the EROS results showed an early IFCOC pattern to deviant preceded by a short standard train, while an OC-late IFC pattern when preceded by a long train. In detecting abstract changes, no EROS match response was observed for deviant preceded with a short train; however, an early IFC-OC- late IFC pattern was observed for deviant preceded with a long train. These results demonstrated 1) the spatial-temporal dynamics in pre-attentive detection of visual changes is similar to that in detecting auditory changes; and 2) activation patterns that was consistent with the prediction violation framework.