Cortical Circuitry Underlying Memory-Guided Sensory Decisions

As we interact with our environment, the features of objects in the visual scene are not consistently present on the retina and sensory cues used to guide visual behavior are not always available. Thus, active observers are faced with a ubiquitous task of comparing sensory stimuli across time and space. The mechanisms underlying such tasks are likely to involve cortical regions sub-serving processing of sensory stimuli, their storage, as well as areas controlling visual attention and decision-making.

Our research program is aimed at examining cortical circuitry underlying successful execution of sensory comparison tasks involving visual motion. We record the activity of MT neurons specialized in motion processing and of neurons in the prefrontal cortex (PFC), an area strongly associated with executive function, sensory working memory and attention. We are particularly interested in the still poorly understood influences of the PFC on sensory cortex. Our experiments are designed to characterize the representation of visual motion in the PFC and to examine the top-down influences it provides to MT during motion discrimination tasks. Our recent work revealed that the majority of the PFC neurons show selectivity for behaviorally relevant motion direction and speed, suggesting its MT origins. We have also characterized the circuitry involved in assigning task relevance to such stimuli and examined memory-related signals its neurons carry. We record spiking activity and local field potentials and measure perceptual thresholds while monkeys compare various features of two sequential stimuli presented within and between different portions of the visual field.

Our current projects include the study of motion representation in the PFC across space aimed at determining the local nature of the bottom-up signals arriving from MTs in both hemispheres. This information has important implications for the way PFC neurons interpret sensory signals appearing in different portions of the visual field and for its top-down influences on the highly retinotopic and stimulus selective MT neurons. Another project is focused on the comparison of neural representation of motion and its location during memory guided discrimination tasks. We also study the behavior of neurons in area MT and their interactions with neurons in the PFC during the same behavioral tasks. To determine the influence the PFC on activity of MT neurons during all components of motion comparison tasks and its contribution to behavioral performance of these tasks we use selective reversible inactivation of regions in the PFC shown to be active during such tasks. The lab has an active ongoing collaboration with Dr. Albert Compte, the head of the Computational Neurobiology Lab at IDIBAPS (Barcelona, Spain), who has been developing biologically plausible computational models of working memory focusing on temporary storage of both spatial and motion information.

Our studies of the way PFC represents and controls sensory signals used during memory-guided sensory tasks have important implications for elucidating the basis of cognitive dysfunction in mental disorders, long associated with deficits in sensory working memory and impaired prefrontal function.