Past events

 Vision is an active sense in which an animal's gaze and pupil shape the content of the retinal image. In the first part of my talk, I will discuss how the viewing strategies of mice align with the neural architecture of their visual system to accomplish an essential visual task: predation. In the second part of my talk, I will compare the hunting behavior of mice to that of a specialized predator, similar in size but distant in evolution, and present our initial insights into the organization of visual information in this animal. Finally, I will present ongoing work indicating that the pupillary reflex arc implements a more complex stimulus-response function than previously thought. I will discuss the underlying neural mechanisms and potential purpose and show conservation from mice to humans. 

We developed an arena (called colloquially the RIFF) for jointly studying behavior and neural activity in freely-behaving rats. The RIFF operates as a state machine, allowing us to implement a large number of different behaviors as Markov Decision Processes and therefore to analyze much of the data within the theoretical framework of reinforcement learning. In the studies I will show here, we recorded neural activity from auditory cortex while rats performed auditory-guided behavior. We observed an intricate interplay between behavior and neural activity that was much richer than we expected.

A plethora of studies have pointed to sensory plasticity in the adult visual system, documenting long-term improvements in perception. Such perceptual learning is enabled by repeated practice, inducing use-dependent plasticity in early visual areas and their readouts. I will discuss results from our lab challenging the fundamental assumption in low-level perceptual learning that only 'practice makes perfect', indicating that brief reactivations of visual memories induce efficient rapid perceptual learning. Utilizing behavioral psychophysics, brain stimulation and neuroimaging, we aim to reveal the neurobehavioral mechanisms by which brief exposure to learned information modulates brain plasticity and supports rapid learning processes. In parallel, we investigate how these learning mechanisms operate across domains, for example by testing the hypothesis that similar inherent mechanisms may also result in maladaptive consequences, when brief reactivations occur spontaneously as intrusive enhanced memories following negative events. Unraveling the mechanisms of this new form of rapid learning could set the foundations to enhance learning in daily life when beneficial, and to downregulate maladaptive consequences of negative memories.

My lab’s focus in recent years has been split between development of ultra-high resolution fMRI at high field and the exploration of more sensitive yet robust methods to find all the salient transients and trends in the signal. High field, high resolution fMRI relies heavily on the acquisition technology and the functional contrast used as well as unique processing approaches that segment, as well as possible, cortical layers for analysis. Our fMRI time series analysis research relies on creative paradigm design in conjunction with tailored processing methods that strike a balance between casting a wide net for potentially informative signals and applying just enough modeling to make sense of the data. Our goal is to use fMRI to see neuronal activity and capture neural correlates of behavior that have previously been elusive to more standard approaches.  Specifically, for our high resolution fMRI work, I will describe experiments demonstrating layer-specific activity in motor, somatosensory, and visual cortex that changes with tasks that modulate the hypothesized input and output cortical communication. In our lab, we perform layer fMRI using a functional contrast called VASO (vascular space occupancy) that is sensitive to blood volume changes in micro vessels - having more specificity than BOLD with only a small tradeoff in sensitivity. Layer fMRI has the potential to provide cortical hierarchy information and communication directionality based on the understanding that feedforward connections terminate predominantly in middle layers and feedback connections terminate in predominantly upper and lower layers. Hence by determining activation location across cortical depth, one can infer whether the activation is feedforward or feedback. I will also demonstrate how the use of resting state connectivity in conjunction with layer fMRI is able to discern such cortical hierarchy in visual areas. Lastly, I will also show examples of applications of layer fMRI in frontal cortex during a working memory task. In addition, I will show our high resolution fMRI work that has allowed us to discern a new digit organizational pattern in motor cortex.  For our time series work, I will show our recent results in using connectivity-based decoding for identifying, in an unsupervised manner, tasks being performed. In addition, I will show an application of naturalistic stimuli and inter subject correlation to characterize personality trait and language skills of individuals. Lastly, changes arousal state during scanning has been viewed as both a confound and opportunity. I demonstrate our effort to further characterize the temporal and spatial signatures of arousal state changes in fMRI time series.