Report

Teaser, summary, work performed and final results

Periodic Reporting for period 1 - MCircuits (Connectivity, plasticity and function of an olfactory memory circuit)

Teaser

Brains accumulate knowledge about the world to inform intelligent behavior using strategies that are different from artificial intelligence. It is thought that this process is mediated by experience-dependent changes in the connectivity between neurons in a network...

Summary

Brains accumulate knowledge about the world to inform intelligent behavior using strategies that are different from artificial intelligence. It is thought that this process is mediated by experience-dependent changes in the connectivity between neurons in a network. Theoretical work proposed that experience-dependent modifications of network structure could be a self-organizing process that results in a representation of relevant structure in the world and permits efficient access to such representations to inform future behaviors. Direct insights into the organization and plasticity of network connectivity in the brain are therefore of key importance to understand the neurobiological basis of intelligence, the dysfunctions of neuronal circuits in disease, and the relationship between biological and artificial intelligence. However, dense reconstructions of connectivity between populations of neurons present major technical challenges, particularly in large brains. We address this challenge by a combination of cutting-edge technologies in a small vertebrate model organism, the zebrafish. We measure neuronal activity across large populations of neurons evoked by sensory stimuli, analyze the modifications of sensory responses during learning, determine the underlying modifications in network connectivity, and further analyze information processing using computational modeling. The results are expected to provide novel and fundamental insights into network mechanisms of memory storage and information processing in the brain.

Work performed

We established the required methodology, computational infrastructure and instrumentation. Experiments in all work packages have been initiated and are proceeding as planned with only minor deviations from the schedule. Progress has been good particularly in the mission-critical parts of the project. The work included the setup of a new microscope for large-volume electron microscopy, experiments to understand the function of inhibition in neuronal networks during learning, and the generation of brain samples to analyze learning-dependent modifications in network connectivity.

Final results

The results are expected to provide fundamental insights into information processing strategies that are directly relevant to inform future developments in artificial intelligence. The interest of computer scientists in the structure and plasticity of biological neuronal networks is extremely high but solid data are very rare. This project is expected to generate this type of data. Moreover, the results will provide a starting point to investigate network-level dysfunctions of the brain associated with neuropsychiatric diseases. Because the study is pioneering both new technologies and new scientific questions, it is thought to open various new directions for future research. Techniques developed during this project are expected to be widely used in future research for a broad spectrum of applications.