Overview
The raw sensory information that enters our nervous system is often chaotic and overwhelming, yet we effortlessly turn these signals into simple and interpretable knowledge about the world. It is our brains, rather than our eyes or ears, that achieve this feat by combining sensory input with environmental context, ongoing behavior, and unique life experiences. The result is a flexible sensory code that allows us to operate at a high level in a broad range of scenarios. Our goal is to probe the structure of this flexible code by inserting ourselves into the middle of the interaction between behavior, experience, and sensation.
We harness the unique properties of the auditory system to attach sounds to typically quiet behaviors and environmental objects, allowing us to establish an animal’s sensory history and alter it at will. Our experiments combine head-fixed and freely moving behaviors, high-density multi-area neural recordings, wireless neural recordings in freely moving animals, neural manipulation, anatomy, and computation. Our work will help us understand central auditory processing in health and disorder while uncovering principles of neural computation that will apply broadly throughout the brain.
Projects
Auditory Thalamocortical Circuits: Understanding the computations performed in the cerebral cortex and how these computations are generated by neural circuit architectures is a central goal of neuroscience research. Rather than operating in isolation, however, most cortical areas make reciprocal connections with one or more thalamic regions. We seek to understand how distributed interactions between the thalamus and cortex underlie important sensory processing computations. This line of research involves characterizing and recording from defined subregions of the auditory thalamus and cortex while animals perform sensory tasks. We are particularly interested in the short time-scale interactions between the higher order thalamus, the primary sensory cortex, and the motor cortex.
Motor-Sensory Predictions: Learning from the past to make predictions about the future is a key computation performed in the central nervous system. One prevelant example is our ability to anticipate the sensory outcomes of our own actions which is critical for an array of behaviors. Our goal is to uncover the specific circuit structures that implement these motor-sensory predictions as well as their impact on sensory encoding. To do this, we create engineered behaviors where we attach sound outcomes to typically quiet mouse behaviors. This allows us to perform large scale recordings from targeted neuron populations while mice experience expected and unexpected outcomes of their actions.
Flexible Sensory Processing: Motor-sensory predictions are just one form of prediction and predictability is just one of many contextual factors that influence the way sensory information is encoded. We know that the brain uses a flexible sensory code to represent incoming information in a way that is tailored to momentary behavioral needs, but we know very little about the structure of this code. Our goal is to capture the dynamic switches between processing modes by observing neural activity as animals experience the same sensory event across a range of behavioral scenarios. To accomplish this, we perform wireless neural recordings while mice live in acoustically enriched home cages that facilitate a range of engineered and naturalistic sensory experiences.
Technology
High Density Electrophysiology: State-of-the-art recording technologies from Neuronexus allow us to record from more than 1000 channels simultaneously distributed across multiple brain areas.
Wireless Electrophysiology: Wireless electrophysiology technology from White Matter LLC enables recording of neural activity while animals explore immersive environments without tethers.
Engineered Motor-Sensory Behaviors: Designing custom closed-loop behavioral software and aparatuses allows for user controlled linking of precise auditory feedback and animal movements.
Acoustically Enriched Arenas: Custom-engineered home cage environments allow animals to engage with more - and increasingly complex - sound-generating behaviors at their own pace and in an environment conducive to learning.
Behavioral Analaysis: Rapidly advancing open-source behavioral analysis tools allow us to perform unbiased and high-throughput analysis of animal behavior in head-fixed and freely moving conditions.
Anatomical and Genetic Tools: The expansive genetic toolkit available in mice allows us to perform many types of neural tracing, neural identification during recording, and neural perturbation.