How do emergent properties develop in the brain in its early stages of development? One of these properties and an essential characteristic of the visual cortex is its ability to recognize moving contours' direction, the so-called orientation selectivity. Measurements of neuronal activity in the living and seeing brain would be required to investigate how this orientation selectivity emerges. However, in the anesthetized animal, orientation selectivity measures are only possible from a few neurons and specific time. To overcome these limitations, we have developed purely optical techniques that combine a surrogate "brain" with an in-silico visual pathway. The in-silico pathway consists of a computer model that mimics retinal and thalamic processing of the visual input and projects the result onto neural networks grown in vitro. These neurons are made photosensitive by expressing genetically light-activated ionic conductivities. Simultaneously, we measure neuronal activity within the network by genetically expressed fluorescence indicators that depend on neuronal activity. This purely optical, minimally invasive system allows the observation of complex neural patterns over up to 21 days. We have succeeded in tracking orientation selectivity in the neural network over these long periods, even though they never experienced actual visual stimuli.
We are currently experimenting with other excitation stimuli as they are "seen" by other animals and comparing the results of in vivo measurements of other research groups. Furthermore, in collaboration with Fred Wolf and his group "Theoretical Neurobiology", our method allows us to validate theoretical models of brain development experimentally.
After the current outbreak of the pandemic, all group efforts have focused on developing affordable but reliable and rapid solutions for the early detection of COVID-19 infections, in collaboration with David Gomez and his group "Systems Medicine Innovation". A prototype is already delivering promising results.