NextGenVis key methods and approaches

Functional Magnetic Resonance Imaging (fMRI): As shown by network partners, state-of-the-art ultra-high resolution 7T and 9.4T MRI scanners –available in the network – have finally enabled describing spatio-temporal activation patterns in the human visual cortex at sub-millimeter resolution. This allows sub-millimeter resolution investigation of distributed cortical networks and laminar activation profiles7. To meet our challenges, we will take advantage of the vast possibilities offered by these new scanner capabilities. Our fellows will be taught the physical backgrounds of the methods they use, yet our training and research activities do not aim at advancing scanner technology and sequences. However, we will closely monitor – and expect to benefit from –further developments in this field, in particular through our many and often direct contacts with groups that focus on this aspect of improving scanner capabilities.

haak1

Eye-movement recording and analysis: Team members have developed specialized software (e.g. Matlab EyelinkToolbox) and pioneered the use of eye-movement coupled fMRI analysis in visual exploration to infer neural, perceptual and cognitive processing. The network has multiple systems for high-quality non-invasive eye- movement recording for use both inside and outside MRI scanners.

eyetracker2

Non-invasive electrophysiology: Team members have extensive expertise in applying these techniques in vision. Clinical approaches include state-of-the-art multifocal flash and Pattern electroretinograms (mfERG, mfPERG), and multifocal Visually Evoked Potentials (mfVEP). They are an indispensable tool for a detailed spatially resolved characterization of various visual pathway pathologies and for objective visual field assessments.

Electro-Magnetic Source Imaging (EMSI): This is an extension of conventional VEPs in which electrical data from high-density electrode arrays are combined with detailed anatomical models produced from MRI scans. EMSI has very high temporal resolution and is therefore complementary to fMRI which has relatively high spatial resolution. Optimal results are achieved when fMRI and EMSI are combined to yield a hybrid technique with high spatiotemporal resolution. In this case fMRI is used to identify individual cortical visual areas while EMSI provides high-resolution electrical time-course data from these areas. This version of the EMSI technique will allow us to examine responses in both normal and abnormal human visual cortex with millisecond temporal resolution.

Population Receptive Field (pRF) modeling: This new class of tools for analyzing brain activity allows a characterization of each recording site according to biologically-inspired computational models using neuroimaging such as fMRI. Recording sites can be described in terms of stimulus representations and connectivity. Team members are at the forefront of the development and application of these tools. The tools initially captured only relatively simple properties of the pRF such as position (retinotopy) and size, but were recently extended to capture suppressive surrounds, compressive non-linearities and their cortical representation, cortico-cortical relationships (connective fields (CF)) and even cognitive representations like attention and numerosity.

prf

Computational modeling: Team members have expertise in modeling many aspects of visual processing (e.g. color vision, crowding, binocular rivalry, motion perception9) and in the wide range of computational methods (for example, Bayesian modeling, neural networks, machine learning and pattern recognition) that will be used to model: a) adaptive processes in perception that operate at different time scales at the level of neuronal populations and networks. (Modeling aims at increasing understanding of the visual system in health and disease, rather than merely reproducing its complexity), b) adaptive software systems informed by our newly acquired knowledge.

Morphometric and connectivity analyses: Team members are experienced in using MRI to map brain structures and connectivity, e.g. to map changes in cortical grey and white matters in relation to eye-diseases11 (e.g. using Voxel-Based Morphometry and Diffusion Tensor imaging (DTI)). Techniques can be used to determine morphometric and connectivity changes in patients.