Catching cognition in the act: new ways to track visual processing, attention and learning in mice

On December 3, 2020 at 11 am (BRT)

Webinário Martha portal

Speaker: Dr. Martha Nari Havenith, Max Planck Research Group Leader - Ernst-Strüngmann Institute for Neuroscience

Moderator: Jean Faber de Abreu, professor at the Department of Neurology and Neurosurgery - Escola Paulista de Medicina (EPM/Unifesp)

We may not be great at multitasking, but our brain is: Neuronal populations (e.g. in visual cortex) constantly represent ‘superimposed’ cognitive processes like perception, attention and learning that are simultaneously ongoing. While these processes have been studied separately, it is unclear how they can be encoded jointly at any given moment. One obstacle to such research is that behavioural tasks generally provide multi-trial estimates of just one cognitive process (e.g. attention) in isolation.

We have developed a behavioural paradigm for mice referred to as the Virtual-Environment Foraging (VEF) task, in which animals navigate towards visual targets while ignoring distractors in a virtual environment. The VEF task features short training times and simultaneously tracks several aspects of behaviour, including visual acuity, cued and sustained attention and rule learning, on a single-trial basis.

This approach yields several new insights: 1) Mice have higher visual acuity than previously shown, distinguishing orientation differences of 5°; 2) Mice can correctly predict new task rules long before executing them correctly; 3) Miniscule reward differences (e.g. of 10 μl) cancue animals to adjust their performance strategy, prioritizing either speed or accuracy; 4) Mice consistently show rhythmical fluctuations in attention, but differ widely in their ability to maintain high-attention states; and 5) During task learning, high-attention states remove ‘careless’ but not ‘honest’ mistakes.

A core goal of neuroscience is to elucidate how neuronal activity generates behaviour. While neuronal recording techniques have greatly improved, behavioural measurements require just as much expansion and refinement. Our work aims to advance that cause.

Lido 3981 vezes Última modificação em Quarta, 08 Mai 2024 13:00