Purpose: Measurement of the oscillating CSF flow in the spinal canal (SC) of healthy volunteers and in patients with post-traumatic syringomyelia (PTS) using an optimized MRI protocol as well as to determine whether stenosis induced velocity changes are detectable using MRI.
Methods: In 68 healthy volunteers quantitative studies of CSF flow in the cervical, thoracic, and lumbar regions were performed. First, an optimized sequence was developed and tested in 19 volunteers using four different flow-encoding velocities (4, 8, 12, 16 cm/s). Secondly, the optimized sequence was employed in 49 volunteers to measure the different CSF patterns in the cervical, thoracic, and lumbar spinal canals (CSC, TSC, LSC). Part three of the study, in which patients with PTS are being examined is still underway. We measured the maximum velocity (cm/s), the pixel area (mm2), and the stroke volume (ml/s). Using a flow model the velocities prior to and after compression with 5 different power levels were measured at the stenosis and at a distance of 70 cm.
Results: A total of 226 dynamic measurements have been performed--so far 76 in the first part (62 = 81.5% evaluable) and 150 in the second part--using the optimized sequence and optimal flow velocities. A flow-encoding sequence of 12 cm/s was found best in the CSC and one of 6 cm/s in the TSC and LSC. The maximum velocity in the CSC was 0.95 cm/s with the flow being directed caudal and 0.38 cm/s with the flow being directed cranial. In the TSC the values were 4.7 cm/s and 1.65 cm/s and in the LSC 0.96 cm/s and 0.59 cm/s. The highest velocities were found at the TSC, which has the smallest diameter compared to the CSC and LSC. In the 4 patients with PTS, the maximum velocities were between 0.09 cm/s and 0.97 cm/s with the flow being directed cranial and between 0.04 cm/s and 1.03 cm/s with the flow being directed caudal. The stroke volumina in the CSC were between 0.1 and 1.23 ml/s (mean: 0.48 ml/s) and 0.2 and 2.45 ml/s (mean: 0.66 ml/s) in the TSC and in the LSC 0.08 ml/s and 0.67 ml/s (mean: 0.29 ml/s). The results of the flow model studies showed an increase of velocity between 2.06 and 4.94 cm/s (mean: 3.31 cm/s) at the stenosis and 1.1 and 1.33 cm/s (mean: 1.23 cm/s) at a distance of 70 cm.
Conclusion: Quantitative measurement of the oscillating CSF flow in the entire spinal canal (SC) is possible using an optimized MRI protocol as well as to detect stenosis induced velocity changes. Due to the high interindividual variability in the data of spinal CSF dynamics, further studies are necessary to collect normal data. The detection of movement of CSF in a post-traumatic spinal cord lesion may alter the therapeutic management.