Seeing the big picture: the function of top-down signals in the visual system
Supervisors:
Dr Matthew Self, School of Psychology & Neuroscience
Prof Lars Muckli, School of Psychology & Neuroscience
Dr Lucy Petro, School of Psychology & Neuroscience
Prof Daniele Faccio, School of Physics & Astronomy
Summary:
A PhD project in the labs of Dr. Self and Prof. Muckli will investigate top-down signals in the visual cortex of mice and humans. In the visual system, top-down connections from higher-visual areas outnumber bottom-up connections from the retina, yet our understanding of their function is limited. Predictive-coding theories suggest that top-down signals contain predictions of sensory inputs which make coding more efficient by subtracting away predictable inputs. This has been investigated with occlusion paradigms, studying top-down neural signals in parts of the visual scene that have been occluded. These signals carry information about the surrounding non-occluded image regions. The candidate will investigate if top-down signals contain predictions of sensory inputs or contextual information. The candidate will be trained in modern, multi-neuron recording techniques in mice (Neuropixel recordings) and high-field laminar fMRI in humans and training in advanced brain-reading approaches. Additionally, optogenetic manipulations in mice will identify the sources of top-down signals. This project will contribute to our fundamental knowledge of the function of top-down signals. As well as multiscale, multispecies interdisciplinary training in empirical brain reading, the PhD student will receive some foundational theoretical knowledge in brain-reading applications, such as brain–wearable technologies for monitoring and decoding neural activity.