CFD of Cerebrospinal Fluid (CSF) in the Spinal Canal
Goal: Investigate fluid dynamics environment of CSF in the spinal canal.

Full 3D transient CFD analysis was performed on the healthy volunteer and two Chiari malformation (CM) patients before and aftere surgery.  Subarachniod space (SAS) geometries (spaces where CSF flows) in the spinal canal and flow waveforms were obtained from MRI scans.  The geometries had an annular structure throughout the whole computational domain.  Rigid walls were assumed.  Flow waveforms were used as inlet BC's with blunt profiles.  To assure the quality of CFD results, the extensions were added.  At the opposite end of inlet BC's, the constant pressure BC's were employed.  Density and viscosity were same as of water.

The results have shown a spectrum of velocity profiles.  Depending on the cardiac cycle position, in larger gaps the blunt velocity were observed, indicating inertia dominated system.  In tighter gaps, parabollic-like profiles were observed with indication of viscous effects.  In transitions from the cardiac cycle peaks (systole and diastole), m-shaped velocity profiles were observed.

In order to asses the differnces between the five cases (healthy and two patients with before or after surgery), pressure drop was extracted for each case.  With same flow waveform as used in inlet BC, it was possible to compute the modulus of longitudinal impedance, which could be known as unsteady resistance to CSF flow.

Images and animations
Healthy volunteer simulation (upper back region)
- Healthy CSF (w-comp velocity) ~1.9 mb -> Reference+Geometry+Waveform
5 different cases (sub-cranial region)
- SAS geometries - CSF flow waveforms
- Healthy CSF (velocity mag.) ~1.4 mb -> scale (m/s)
- CSF flow with Chiari malformation (velocity mag. - m/s) ~1.9 mb -> scale (m/s)

Unsteady Resistance -  Longitudinal Impedance
Goal: Obtain a parameter for CSF resistance with computed pressure drop and flow.
 

Images and animations
5 case comparison of CFD calculations (data from above)
- Modulus of longitudinal impedance (5 cases)

- Integrated longitudinal impedance modulus (5 cases)

Phantom Model of the Cerebrospinal Fluid (CSF) System with Chiari Malformation (CM)
Goal: Experimentally simulate the CSF flow in the restricted spinal canal with Chiari malformation

This projecty used MRI utilities to investigate hydrodynamic environment of CSF in the sub-cranial subarachnoid space (SAS) with CM.  A model was constructed to simulate fluid dynamics of CSF in SAS with CM. This model will allow investigation of detailed pressure and velocity using laser Doppler anemometry and pressure transducers. The MRI data from patient and the flow model were compared.

This study was directed toward better understanding of the hydrodynamics within the spinal canal and syrinx.  This study demonstrated that the physical model reproduced in-vivo hydrodynamic environment in the CM/SM patient.  The characterization of the hydrodynamic environment using modeling may lead to greater understanding of pathogenesis of craniospinal disorders such as CM and SM.

Images and  animations

- 3D model of CM reconstructed in Mimics -> Animation ~1.6mb

- Sylgard model of CM

- Peak systole velocities compared between patient and Sylgard model

- Animation of flow in patient and Sylgard model at C2 ~0.9mb

- Animation of flow at three location within the Sylgard model ~0.8mb

MR Measurement of Pulse Wave Velocity in the Spinal Canal

Goal: Compute pulse wave velocity from MR images

Non-invasive measurement of pulse wave velocity (PWV) in the CSF system is of interest as a potential indicator of subarachnoid space pressure and compliance, both of which play a role in the development of craniospinal diseases.  PWV measurements using a novel MR technique that acquires unsteady velocity measurements during the cardiac cycle with a time interval <10 ms are presented.  Axial CSF velocity measurements were obtained in the sagittal plane of the cervical spinal region on three patients without cranio-spinal disorders.  PWV was estimated by using the time shift identified by the maximum velocity and maximum temporal velocity gradient during the cardiac cycle.  Only those that were statistically significant were considered valid. The methodology represents a new technique that can be used to measure PWV in the spinal canal non-invasively.

Images
- Example of one case
- Maximum velocity
- Maximum temporal velocity gradient during the cardiac cycle

MRView - Matlab GUI
Goal: Compute flow waveforms from phase-contrast MRI.

This in-house matlab code generates flow waveforms form PCMR images.  Also it has capability to visualize and animate the CSF and blood flows.

Images and animations
- MRView - Matlab GUI
- Example of CSF flow in SAS (velocity - mm/s) ~0.6 mb
- Example of CSF flow in SAS and Syrinx (velocity - mm/s) ~0.3 mb