Octavio Ruiz


The excellent work of many scientists has taught us that different areas of the brain are specialized in different aspects of our visual experience (for example: detection of borders between objects, colors, motion, and face recognition).

Less is known, however, about how the brain processes the signals sent by our retinas to the brain: how features are extracted, how they are combined to produce a percept, how the brain copes with the immense complexity and variety of natural scenes, and what is the role of individual neurons in that process.

My work attempts to extend our knowledge of the primate visual system by combining the study of vision in natural conditions, with novel methods to manipulate neuronal activity and observing the consequences.

Talks:  UCSD Dpt. of Psychology, 2014 may 06

Natural vision

Much has been discovered using simple images to probe the visual system. To examine whether those basic mechanisms hold under the conditions of daily life, my colleagues and I are studying the visual system under increasingly-complex naturalistic conditions. When differences relative to simple conditions are found, we study the origin of those differences, and try to incorporate them into our models of the visual system.

For example, we showed that scene complexity and eye movements (saccades) delay and reduce primary visual cortex responses, but transiently increase orientation selectivity. Our report can be found here.

My collaborators (T.Lii, M.Paradiso) and I then investigated whether these effects have a counterpart in human perception. Our data shows that, during normal vision, our visual contrast sensitivity is indeed reduced relative to vision without eye movements (article in preparation).

Manipulation of neurons in vivo

To understand the role of individual neurons in natural vision my work aims to manipulate neural activity of behaving animals and observe the consequences in the rest of the brain and in behavior.

A recently developed tool, optogenetics, uses molecular-genetics to make particular neurons sensitive to light. Then, during an experiment, light is shined directly on the brain to activate or silence those neurons with high spatial (1 mm or less) and temporal precision (milliseconds). An overview of this approach can be found here.

This powerful technique is well established in rodents. In the last few years, we have developed a new system to use optogenetics in NHPs (a collaboration initiated by G.Stoner, T.Albright, E.Callaway, and me, at the Salk Institute, with A.Roe and B.Lustig at Vanderbilt University, and later joined by J.Nassi and J.Reynolds at the Salk). Our system enables for neural manipulations and simultaneous measures of neural activity with two different, complementary techniques: optical-imaging (activity of hundreds or thousands of neurons in a cortical region) and electrophysiology (one or two neurons at a time, but with very high temporal resolution). Our system is described here (or here).

Both lines of work (naturalistic vision and manipulation of neural activity) are separately evidencing important factors present in our visual system that reduced experiments cannot unveil. My next goal is to combine the two approaches to extract information about the algorithms operating in our brain that allow us to see the world.

Long-term acquisition and integration of multiple measures of cortical function and structure (work in progress)

This project will collect multiple-modality data from a ~15 mm x 15 mm x 1.5 mm volume of brain cortex, repeatedly, for an extended period of time (~ 1 year), under different brain states, behaviors, sensory inputs, and optogenetic manipulations.

The data will be registered to a 3-dimensional magnetic-resonance-imaging reconstruction of the subject's brain. Our software resource (to be developed) will allow us to mine the data and answer functional questions about the circuitry of that particular region of cortex.

A description of this project can be found here.

Contact (scientific matters):  oruiz at salk dot edu


Other than Science