Jan Robert Kroger MD, Presenter: Nothing to Disclose

SURAJ THYAGARAJ, Abstract Co-Author: Nothing to Disclose

Daniel Giese, Abstract Co-Author: Nothing to Disclose

Dennis Hedderich MD, Abstract Co-Author: Nothing to Disclose

Richard Lukas Clemens Uwe Morsdorf-Schulte, Abstract Co-Author: Nothing to Disclose

David Christian Maintz MD, Abstract Co-Author: Nothing to Disclose

Theresia Yiallourou, Abstract Co-Author: Nothing to Disclose

Soroush Heidari Pahlavian, Abstract Co-Author: Nothing to Disclose

Alexander Christian Bunck, Abstract Co-Author: Nothing to Disclose

Bryn A Martin PhD, Abstract Co-Author: Nothing to Disclose

The mechanisms of cerebrospinal fluid (CSF) hydrodynamics in the pathophysiology of Chiari malformation Type 1 (CMI) are still poorly understood. The aim of this study was to reverse-engineer 3D printed models of the subarachnoid space (SAS) at the craniocervical junction, based on subject-specific MR measurements, to help understand the accuracy of 4D-phase-contrast-(PC)-MRI and the hydrodynamics in CMI.

4D-PC-MRI and high-resolution T2-weighted MR-images were obtained for a CMI patient and healthy control. Four subject-specific 3D-printed models of the subarachnoid space near the craniocervical junction were constructed based on the in-vivo images, two with idealized nerve roots. A pulsatile computer-controlled pump was constructed to produce subject-specific flow waveforms. The four in-vitro models were scanned by 4D-PC-MRI and peak velocities were compared along the cervical spine for the in-vivo and in-vitro measurements.

For the healthy volunteer, in-vitro and in-vivo flow characteristics were similar. Peak CSF flow velocities correlated with the area of the SAS in all models (r=0.6; p<0.001) and in the healthy volunteer (r= 0.7; p<0.05) but not in the CMI patient (p>0.05). For the CMI patient, in-vivo and in-vitro velocities had poor agreement, particular near the foramen magnum. At this region, in-vivo flow patterns in the CMI patient showed unilateral dominated flow jets and elevated flow velocities that were not present in the corresponding in-vitro models. The in-vitro models with nerve roots showed elevated flow velocities compared to the models without nerve roots. Flow distribution along the cervical spine was similar for the models with and without nerve roots with localized flow disturbances surrounding the nerve roots.

Quantification of the CSF flow field by 4D-PC-MRI showed good agreement with in-vivo flow characteristics in the healthy case and poor agreement for the CMI patient. These differences demonstrate that a replication of static morphology is insufficient to explain the alterations in CSF dynamics seen in CMI patients and that neural tissue motion and/or a systematic error in the 3D model geometry reconstruction could be an important factor.

The simulation of CSF hydrodynamics in our controlled set-up promotes a better understanding of the crucial variables causing the characteristic alterations in CSF dynamics seen in patients with CMI.

Kroger, J, THYAGARAJ, S, Giese, D, Hedderich, D, Morsdorf-Schulte, R, Maintz, D, Yiallourou, T, Heidari Pahlavian, S, Bunck, A, Martin, B, 4D-phase-Contrast Evaluation of Cerebrospinal Fluid Dynamics in a Rigid-wall 3D Printed in-vitro Model of Chiari I Malformation with Idealized Spinal Cord Nerve Roots.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14007851.html