Super-resolution assessment of relative pressure in intracranial atherosclerosis using ML-enhanced 4D Flow MRI

In vivo measurements of intracranial pressure differences using 4D flow are useful to assess health risks associated with intracranial atherosclerotic disease (ICAD). In the clinical routine, approximations of the Navier-Stokes equation are used to derive relative pressure values, which can be inaccurate in intracranial vessels. Recently we presented an algorithm using a virtual work-energy formulation(vWERP). While this technique yields more accurate pressure estimations in intracranial settings, still systematic errors due to insufficient spatial resolution were observed. Here, we apply artificial intelligence-based super resolution for more accurate assessments of pressure values near the stenosis in a cohort of ICAD patients.