Abstract
The human visual system has limited processing capacity. Both behaviorally relevant information and behaviorally irrelevant information compete for representations in this capacity-limited system. It has been further suggested that distractors that are motivationally significant have a stronger distracting influence. We examined the neural underpinnings of these ideas by recording fMRI from participants who detected instances of coherent motion in a random dot kinematogram (RDK) overlayed on three categories of affective images from the International Affective Picture System (IAPS): pleasant (erotic couples), neutral (workplace people), and unpleasant (bodily mutilations). Behaviorally, we found that there was no significant difference in the subject’s ability to detect coherent motion for the three classes of distractors. Applying machine learning (support vector machine) to BOLD responses in two important areas of visual motion processing: ventral visual cortex (VVC) and MT cortex, we found that the decoding accuracy of both pleasant-vs-neutral distractors and unpleasant-vs-neutral distractors was above chance level in VVC and MT and importantly, the distractor decoding accuracy in VVC and MT was negatively associated with the subject’s ability to detect coherent motion. These results suggest that emotional and neutral distractors have similar influences on the processing of the attended motion stimuli and stronger neural representations of distractors in the motion-processing areas of the brain reduce the ability to detect coherent motion in the RDK task.