A dissertation submitted to The University of Manchester for the degree of Master of Science in the Faculty of Science and Engineering (2019).
Within the SpiNNaker group working on neuromorphic hardware and neuroscience tools, I investigated spike-based mechanisms capable of exhibiting spatially periodic firing patterns analogues to grid cell activity in the entorhinal cortex. Grid cells are believed to play a key role in spatial navigation in the mammalian brain. The grid cell model presented in this project is based on continuous attractor neural network (CANN) dynamics with recurrently inhibitory connections, designed to run on the SpiNNaker v1 neuromorphic platform.
This project discusses significant observations in the grid cell model’s activity, showing that while CANN grid cell models exhibit the characteristic features of grid cells, a spiking neuron model with biologically-plausible parameters requires further exploration and tuning. Further insight into the biological mechanisms supporting spatial navigation will result in significant advancements for adaptive robotic navigation, autonomous exploration and the efficient representations of space.
links: