Our work aims to investigate the properties of neuronal circuits and their fundamental computations. Since sensory information processing always takes place in the context of behavioral goals we are particularly interested in studying how neuronal circuit function is modulated by cognitive factors, such as expectations, attention or decision-making. To address these questions we measure the activity of local neuronal populations using multi-electrode recordings and macroscopic optical imaging in the visual system of awake, behaving rodents.
Model Organism
Our model system for studying visual information processing is the rodent. While the rodent has a more simple visual system compared to the classic animal models for vision, cats and primates, basic response properties of neurons are preserved, like simple/complex receptive field structures and selectivity for orientation and direction. In the rodent, the visual cortical system consists of several different areas which are located on the surface of the brain, making them readily accessible for the implantation of electrodes and imaging. Additionally, in rodents powerful genetic tools exist for the analysis and perturbation of neuronal circuits.
Methods
We deliver controlled visual stimuli while the animal is placed on a spherical treadmill or freely moving in a choice arena. The animals are trained to simply view visual stimuli or perform visual tasks during recording of neuronal activity. We employ infrared eye tracking techniques to monitor eye movements during the presentation of visual stimuli.
We use multisite silicon probes to record neuronal activity across the layers of the visual cortex and optical imaging techniques to measure population activity across the surface of the visual cortex. Silicon probes offer the opportunity of assessing local neuronal circuitry by relating recorded activity to the cortical layer. Optical imaging can serve as a guide for targeted electrode placement and can reveal the combined spatio-temporal pattern population activity in output layers. Advanced data analysis methods are employed for the identification of single units and processing of multidimensional time-series. In addition, we exploit optogenetic tools to perturb and manipulate neuronal activity and behavior.