Large-Scale In-Silico CSF Investigations within the Optic Nerve

Studying cerebrospinal fluid (CSF) flow within the human optic nerve is crucial for understanding conditions such as papilledema and glaucoma. We discuss large-scale investigations combining supercomputing with Terabyte-sized synchrotron-radiation microcomputed tomography 3D imaging.

Our study focuses on the quantitative response of CSF pressure gradients and mass transfer relative to morphological alterations in the optic nerve's subarachnoid space (ONSAS) microstructure. This investigation reveals the key role of ONSAS trabeculae. These findings pave the way for future exascale investigations, promising unprecedented insights into fundamental physiology and pathophysiology of vision.

In this talk, we report on our experience in coupling pseudospectral schemes with immersed boundary methods to deal with the complex ONSAS geometry. Deeply rooted in digital filter design, pseudospectral schemes offer exceptional advantages in accuracy, numerical dissipation, and supercomputing readiness. These attributes make them ideal for high-quality direct numerical simulations (DNS) of a broad range of flow problems, as well as for effective processing of large-scale images, including signal registration, stitching, and signal enhancement.