Computational framework for stimulus efficacy in pRF mapping

In the human visual system, light impulses are converted into nerve impulses in the retina, which are then processed in the visual cortex. Retinotopy is one of the most fascinating features of this visual system. It refers to the fact that information in adjacent areas of the visual field is processed on adjacent regions in the visual cortex. Thus, each point on the visual cortex is associated with a specific section in the visual field. Functional magnetic resonance imaging (fMRI), as a completely non-invasive technique, is ideally suited to study nerve cell activity in the visual cortex. Functional MRI takes is based on the different magnetic properties of oxygenated and deoxygenated blood in order to obtain maps of brain activity. Using periodically changing visual stimulation patterns, fMRI can be used to study the retinotopic features of the visual system. The aim of this project is to develop new stimulation patterns whose effectiveness is significantly higher than the standard patterns used so far. This will drastically reduce the acquisition time in fMRI and thus improve the applicability of retinotopic measurements, particularly in clinical settings. In this project, the expected fMRI data of the visual cortex are simulated and the retinotopic maps are calculated from them. By repeating the simulations thousands of times, precision and accuracy can be determined under different conditions. Since these simulations involve high computational effort, the algorithms used must be implemented on high-performance computing (HPC) platforms to achieve acceptable computation times. Only then is it possible to test a variety of stimulus configurations. Accompanying this, the stimulation patterns are evaluated using fMRI measurements in healthy subjects to ensure that the measures of effectiveness used in the simulations also correspond to the experimental results. The computational framework developed in this project will for the first time offer the possibility to calculate the effectiveness of arbitrary stimulus configurations and thus to obtain stimuli optimized for a specific application. This will not only significantly improve the quality of retinotopic maps in basic science, but also clinical applications, as stimulus patterns optimized for the specific type of an ocular disease can be generated. The resulting more precise characterization of the disease can improve diagnosis and therapy of affected patients.