Rick Born, MD ’88
Professor of Neurobiology
From smelling coffee in the morning to seeing and recognizing a colleague at work, our senses are the primary means by which we gather information about the external world. The Born Lab studies how different regions of the cerebral cortex endow us with the ability to see.
Ordinarily we think of visual processing as proceeding in stages, beginning with the eye, then the visual thalamus, followed by the primary visual cortex (V1) and a whole host of higher-order visual areas in other areas of the brain called V2 and V3. This so-called “feedforward” direction of information flow is clearly critical for visual perception. However, somewhat surprisingly, information also flows in the opposite, or “feedback,” direction, and the role these connections play in vision is a great mystery. Our lab has focused on how the largest source of feedback to V1, that from V2, influences the computations performed there.
In earlier work, we discovered that feedback from V2 modulates the influence of the visual context—or background—manifest as a reduction in the strength of “surround suppression” when feedback was temporarily inactivated. Interestingly, our results were strikingly similar to those obtained when a particular subset of neurons, known as Martinotti cells, were inactivated within V1, suggesting that feedback might interact with this local circuitry.
This potential relationship was further strengthened by more recent work, spearheaded by postdoctoral fellow Till Hartmann, PhD, in which we studied the effect of feedback on brain rhythms in V1. Brain rhythms refer to specific patterns of neuronal activity, generally classified by the frequency at which they oscillate. Such oscillations have long been studied as proxies of circuit output (many will be familiar with “EEG” measurements used by psychiatrists and neurologists to measure brain activity non-invasively), but their precise role in cortical computation remains controversial. Our interest was initially piqued by reports that oscillations with different frequency ranges were signatures of different directions of information flow: the faster gamma oscillation was purported to be a signature of feedforward processing, whereas slower rhythms, such as alpha and beta, were thought to reflect feedback.
To our surprise, when we transiently inactivated cortico-cortical feedback from V2, it was the gamma oscillations that were most strongly reduced, and this despite a net increase in the spike rates of neurons we recorded simultaneously. This finding challenges the notion that gamma rhythms reflect feedforward processing, as well as the dogma that they are generated purely locally. The result also further strengthened the possible interaction of feedback with Martinotti cells, as a similar reduction in gamma oscillations was seen with their inactivation. Taken together, our findings have implicated specific circuit mechanisms through which feedback might exert its effects. This is an important step toward a better understanding of the role of cortico-cortical feedback in visual perception and the neural circuits that ultimately lead to our ability to see.