Role of fixational eye movements and saccades in processing spatial information in V1-V2-V4 networks

点击翻译按钮获取中文摘要

This project focuses on cortical mechanisms in areas V1, V2, and V4 (V1-2-4) underlying the information- processing roles of saccades and fixational eye movements (FxEMs: microsaccades and drift) in vision. Both saccades and microsaccades are followed by drift. Sequences of saccade/drift and microsaccade/drift cycles, both of which we will refer to as saccade/drift cycles (SDC’s), are believed to be critical to gathering information about visual scenes and the objects within them. We will record single-unit and local field potential (LFP) activity from the foveal projection in areas V1-2-4 simultaneously while monkeys perform match-to-sample tasks for shape or texture, with shapes filled with texture, to understand the role of SDC’s in cortical processing during active vision. Through the use of gaze-stabilization, we will be able to disrupt the retinal input to the cortex during the task to probe selectively how SDC’s control processing of shape and texture in early visual areas. In active vision, the SDC maps spatial information into temporal patterns of neural activity. Recent modeling and psychophysical work have determined that the initial transient phase of the cycle encodes coarse features, while the later, more prolonged period of drift is critical for extracting fine details. We are interested in how the transition from coarse to fine processing in the SDC is implemented in local and inter-areal cortical networks. We hypothesize that processing in these networks during the SDC begins with a phase of transient feedforward activity followed by a longer phase of recurrent and re-entrant activity that coincides with gamma oscillations in the networks. We hypothesize that shape is captured in the initial phase of the SDC and finer features of the shape’s border and the region within the shape by the recurrent phase of processing during drift. We also suggest that the cortex uses sequences of SDC’s to accumulate information and that the organization of networks within the cortex, in terms of their hierarchy and temporal frequency band for communication, depends on whether the sequence is dominated by saccades and drift (looking) or by microsaccades and drift (fixating). In AIM 1, we focus on activity phase-locked to the SDC where the match to sample for shape or texture is fixed for an entire day’s session. The SDC’s in this task will be dominated by microsaccades/drifts. We will ask if accurate matches for shape coincide with stronger SDC transients in the network and if accurate matches for texture produce robust gamma coherence that emerges later in the SDC. In AIM 2, we investigate sequences of SDC’s in periods of looking and fixating for a match-to-sample task where the monkey is given a cue on every trial for the correct match of shape or texture. Information is gained across the trial through saccades and FxEMs, and we ask how the dynamics of V1-2-4 network activity, phase-locked to these sequences, reflect performance and task. Gaze-stabilization will be used in both AIMs to impact cortical processing of the sample stimulus. How the SDC’s modulate the causal relationships between cortical areas and the dynamics of those interactions will be examined through a new method of analysis, frequency-extracted hierarchical decomposition.