Active projects of the Fitzgerald Lab include:
Brain-scale computation: One can explore brain-scale computation in larval zebrafish by imaging nearly all its neurons at cellular resolution. We are collaborating with several zebrafish labs to build realistic neural circuit models of sensorimotor transformations underlying ethological behaviors. We use these experimental and modeling results to ground our thinking about principles that might govern the large-scale organization of the brain.
Dynamics of hippocampal memory: Modern imaging techniques reveal that memory circuits are dynamic at the synaptic, neuronal, and network levels. We are interested in understanding how the dynamics of memory emerge from multiscale dynamics of brain circuitry. We are currently investigating how plasticity and learning in the hippocampus and neocortex determine the nature and time course of long-term memory.
Artificial neural networks: Neural network functions emerge from the activity of many interacting neurons. We are interested in uncovering mathematically rigorous links between structure, dynamics, representation, learning, and computation in artificial neural networks. These links can provide novel hypotheses for neuroscience research, and our theoretical understanding of neural networks informs our approach to biological circuits.
Visual computation in natural environments: Evolution selects for sensory systems that perform well in natural environments. We study how natural scene statistics affect the visual motion estimation strategies employed by flies, larval zebrafish, and humans. We take a comparative approach to gain general insights into how the mechanistic details of brain circuitry can embody and/or obscure fundamental computational principles.
Physical limits of measurements: Progress in neuroscience is often driven by technologies that enable new types of measurement, yet each technique can only be pushed so far. We are interested in quantifying how the fundamental laws of physics limit neurobiological measurements. These limits help clarify the proper interpretation of new data and define rigorous design criterion for future innovation.