What we learn about the visual system by studying non-human primates: Past, present and future

Organizers: Rich Krauzlis1, Michele Basso2; 1National Eye Institute, 2Brain Research Institute, UCLA
Presenters: Ziad Hafed, Farran Briggs, Jude Mitchell, Marlene Cohen, Nicole Rust, Bevil Conway

< Back to 2021 Symposia

The symposium includes six highly regarded mid-career and junior investigators (Ziad Hafed, Farran Briggs, Jude Mitchell, Marlene Cohen, Nicole Rust, Bevil Conway) who use NHPs to study a range of topics (e.g., attention, eye movements, object and color perception) of interest to the VSS membership. Ziad Hafed will review how an observation about fixations made about twenty years led to a much deeper understanding of the causes and implications of the correlation between attention and microsaccades. Subsequent research also illuminates highly intriguing new questions about the integration of early visual processing capabilities in late motor control brainstem structures. These issues highlight the importance of approaching vision science from an active perspective incorporating the interplay between vision and eye movements. Farran Briggs will describe a series of approaches using traditional and modern tools to explore how cortical feedback influences early visual processing. Transformations in visual signals traversing the feedforward retino-geniculo-cortical pathways are well understood, but the contribution of corticogeniculate feedback to visual perception is less clear. Through examinations of the morphology, physiology and function of corticogeniculate neurons, a new hypothesis emerges in which corticogeniculate feedback regulates the timing and precision of feedforward visual signal transmission. Jude Mitchell will discuss how visual tasks have evolved over the past twenty years. Fixation paradigms have added much to our understanding but sidestep the dynamics of natural vision in which peripheral targets are selected (pre-saccadic attention) and brought to the fovea for detailed inspection (trans-saccadic integration). Jude will describe the application of free-viewing paradigms in the marmoset monkey to examine processing in visual cortex and the dynamics of selective attention across the saccade-to-fixation cycle. Marlene Cohen will describe insights in understanding populations of neurons. Twenty years ago, most NHP work focused on the activity of single neurons and relatively simple stimuli and behaviors. It is now possible to record from many neurons in multiple brain areas while monkeys make judgments about a variety of stimulus properties. Marlene will describe recent work showing that these complex data sets can reveal strikingly simple relationships between neuronal populations and visual perception. Nicole Rust will discuss how deep artificial neural networks (DANNs) have advanced our understanding of visual function. Work over the past twenty years has demonstrated striking parallels between DANNs trained to categorize objects and the functional organization of the primate ventral visual pathway. Nicole will describe work showing that DANNs serve not only as models for object identification, but also extend to other visual behaviors including ‘image memorability’ – why some images are better remembered than others. Bevil Conway will discuss how functional MRI in NHPs has advanced our understanding of the ventral visual pathway. Combining fMRI with neurophysiology has facilitated the systematic study of extrastriate cortex, guided targeted recordings from neurons in functionally identified patches of cortex, and provided direct comparisons of brain activity in humans and monkeys. This work underscores the importance of understanding how functionally identified populations of neurons interact to enable perception of colors, objects, places and faces.

Presentations

A vision for orienting in subcortical oculomotor control areas

Ziad Hafed1; 1Eberhard Karls Universität Tübingen

Ziad Hafed will review how an observation about fixational eye movements, first described about 20 years ago, led to a series of studies exploring the brain circuits that control both attention and saccades. We now have a much deeper understanding of the underlying roots and implications of the correlation between attention and microsaccades, the role of subcortical and cortical visual structures in this process, and the importance of approaching vision from an active, rather than passive, perspective.

The role of corticogeniculate feedback in visual perception

Farran Briggs1; 1University of Rochester

Farran Briggs will describe a series of approaches using traditional and modern tools to explore how cortical feedback influences early visual processing. Transformations in visual signals traversing the feedforward retino-geniculo-cortical pathways are well understood, but the contribution of corticogeniculate feedback to visual perception is less clear. Through examinations of the morphology, physiology and function of corticogeniculate neurons, a new hypothesis emerges in which corticogeniculate feedback regulates the timing and precision of feedforward visual signal transmission.

Neural circuits for pre-saccadic attention in the marmoset monkey

Jude Mitchell1; 1University of Rochester

Jude Mitchell will describe the role of different classes of neurons in visual cortex. Over the past twenty years, there have been major advances towards manipulating and tagging different neuronal classes, and new molecular and recording techniques that distinguish cell class are now becoming available for use in NHPs. Jude will describe the application of these approaches in the marmoset monkey to understand how eye movements modulate early sensory processing as a function of cell class and cortical layer.

Multi-neuron approaches to studying visual perception and decision-making

Marlene Cohen1; 1University of Pittsburgh

Marlene Cohen will describe insights in understanding populations of neurons. Twenty years ago, most NHP work focused on the activity of single neurons and relatively simple stimuli and behaviors. It is now possible to record from many neurons in multiple brain areas while monkeys make judgments about a variety of stimulus properties. Marlene will describe recent work showing that these complex data sets can reveal strikingly simple relationships between neuronal populations and visual perception.

Deep artificial neural networks as models of vision and visual memory

Nicole Rust1; 1University of Pennsylvania

Nicole Rust will discuss the contributions of deep artificial neural networks (DANNs) to our understanding of visual cortical function. Work over the past twenty years has demonstrated striking parallels between DANNs trained to categorize objects and the functional organization of the primate ventral visual pathway. Nicole will describe recent work showing that DANNs serve not only as models for object identification, but also extend to other visual behaviors including ‘image memorability’, or the variation with which some images are better remembered than others.

Parallel multi-stage processing of inferior temporal cortex: faces, objects, colors and places

Bevil Conway1; 1National Eye Institute

Bevil Conway will discuss how functional MRI in NHPs has advanced our understanding of the ventral visual pathway. Combining fMRI with neurophysiology has facilitated the systematic study of extrastriate cortex, guided targeted recordings from neurons in functionally identified patches of cortex, and provided a direct comparison of brain activity in humans and monkeys. This work underscores the importance of understanding how functionally identified populations of neurons interact to enable perception of colors, objects, places and faces.

< Back to 2021 Symposia