Output details
15 - General Engineering
Imperial College London
A finite rate of innovation algorithm for fast and accurate spike detection from two-photon calcium imaging.
This was the first paper to make use of sparse sampling theory to detect action-potential-induced calcium transients in brain imaging data obtained by multiphoton microscopy. Solving this problem allows the activity of large numbers of neurons to be monitored simultaneously. Detection accuracy using prior approaches has been severely affected by signal-to-noise ratio and bandwidth constraints; our new approach substantially outperforms previous methods, meaning that much larger neural populations can now be monitored. The work, published in the leading journal in neural engineering (IOP Science), led to international speaking invitations (e.g. EPFL June 2013, Paris 2013).