Any form of neural information processing, including our thoughts, perceptions, and plans for the future, depends on spatiotemporal patterns of electrical activity that propagate through neural circuits. The spatial component is simple to read out by taking a snapshot of neural activity at a given point in time. Temporal patterns, on the other hand, are more challenging to interpret, especially in systems that consist of millions or even billions of nerve cells.
We approach this problem by studying temporal signal processing in the brain of the fruit fly, which contains only approximately 150,000 neurons, whose connectivity and whose genetic identities are well-defined. This makes mechanistic ideas precise and testable. Like our brains, the fly brain processes temporal signals over at least nine orders of magnitude ranging from action potentials that last only milliseconds up to circadian rhythms and beyond. While the biophysical mechanisms that act at both ends of this spectrum are well characterised, little is known about the processes that operate at intermediate time scales from hundreds of milliseconds up to minutes.
Our work is based on the premise that nervous systems across species employ a common repertoire of circuit architectures to delay, accumulate, and store signals over this time range. Bridging biophysics and behaviour, we investigate three related brain functions that unfold over milliseconds up to minutes to uncover principles of how signals are kept in working memory: motion vision, perceptual decision-making, and distance estimation.
