The site of the Media Sphera Publishers contains materials intended solely for healthcare professionals.
By closing this message, you confirm that you are a certified medical professional or a student of a medical educational institution.
Application of a modified Time-SLIP MRI sequence for visualization of cerebrospinal fluid movement in the cerebral aqueduct and cervical spinal canal
Journal: Burdenko's Journal of Neurosurgery. 2019;83(6): 64‑71
Read: 5340 times
To cite this article:
Development of techniques for non-invasive imaging of cerebrospinal fluid flow is an urgent objective. There are several MRI techniques to obtain qualitative and quantitative data on cerebrospinal fluid flow. There is evidence of the use of 4D velocity mapping (4D-VM) [1]. Phase-contrast MRI (PC-MRI) is the “gold standard” for imaging and quantification of cerebrospinal fluid flow [2—4]. Time-SLIP (Time-Spatial Labeling Inversion Pulse) is a method of Arterial Spin Labeling (ASL). This technique was originally developed for non-contrast MR angiography [5, 6]. Time-SLIP MRI may also be used to visualize flow of other fluids: pancreatic secretion [7], saliva [8] and cerebrospinal fluid [9]. This technique may be used to analyze physiology of cerebrospinal fluid flow [10—12], as well as in clinical practice for Chiari-I malformation [13], external hydrocephalus [14] and syringomyelia [15]. PC-MRI gives quantitative information about cerebrospinal fluid flow velocity, requires much time and mandatory cardiac synchronization. Unlike this, Time-SLIP is a convenient method for cerebrospinal fluid flow imaging. The technique is valuable for analysis of qualitative and quantitative data, for example, length of motion (LOM). It is the distance from tagging zone edge to the last pixel with significant contrast enhancement [13]. Cardiac synchronization is implied, that is associated with certain drawbacks (low accuracy in arrhythmias, photoplethysmograph errors, need to use electrodes). Modification of this technique with exclusion of cardiac synchronization would simplify its routine application for imaging of impaired cerebrospinal fluid flow, for example, in diagnosis of hydrocephalus, after neurosurgical interventions with need for cerebrospinal fluid flow measurement.
To evaluate the value of Time-SLIP MRI for cerebrospinal fluid flow imaging in cerebral aqueduct and cervical spinal canal by registering length of motion (LOM) in a phantom and in patients.
All examinations were made using 1.5 T MRI scanner and standard 4-channel radiofrequency coil. The following parameters were used for Time-SLIP mode: TR=8500 ms; TEeff=80ms; slice thickness 5.0 mm; tag-zone width — 30 mm; number of excitations (NEX) — 7; FOV 26×26 cm; MA 128×256; Half fourier; flip angle 90°; black blood time interval (BBTI) =2000/3000ms, corresponds to inversion time; SPEEDER=2; TA 2:16 min. There was a slice in sagittal plane through cerebral aqueduct.
Cerebrospinal fluid imaging in Time-SLIP mode is that tissues within imaging area have contrast enhancement similar to that in FLAIR mode (fluid-attenuated inversion recovery) and tagging area has a T2-weighted contrast enhancement. Thus, cerebrospinal fluid is hyperintense in relation to surrounding tissues. If spin movement from the tagging area occurred within the BBTI and hyperintense area is determined, low signal zone will be visualized as soon as cerebrospinal fluid enters the tagging zone.
Cardiac synchronization was simulated using a photoplethysmograph from an external device that supplies a light signal to the sensor with a frequency of 160 flashes per minute (over patient’s heart rate). Great number of excitations (NEX=7) was used to obtain the largest possible number of CSF coordinates.
A phantom was developed on the basis of dynamic phantom for flow modeling to analyze the accuracy of this technique [17]. The phantom consisted of a freely rotating pulley with a diameter of 100 mm and polyvinyl chloride tube with a diameter of 4 mm filled with distilled water for CSF simulation and fixed on this pulley (Fig. 1a—d).
Clinical survey enrolled 9 healthy volunteers and 12 patients without cerebrospinal fluid flow impairment (9 men and 12 women). Median age was 45 years (min 24 years, max 84 years). Time-SLIP technique was additionally applied in one patient to control the effectiveness of the 3rd ventricle ventriculostomy.
Measurements were made using the Myrian multimodal console. Regions of interest were cerebral aqueduct and cerebrospinal fluid in C1—C2 spinal canal. Statistical analysis of the images was performed using Statistica 8.0 software package (StatSoft Inc., USA). Non-parametric statistical methods were used due to small sample size and heterogeneous group (Mann-Whitney U-test).
In experimental study, (LOM) was determined for 4 different disk oscillation frequencies. These values corresponded to HR 51, 54, 60, 93. It was shown that the recorded length of motion is well correlated for BBTI 2000 and 3000. No significant differences were obtained (Fig. 1d, e).
Three fixed amplitudes of oscillations were also chosen (1.5; 2.3; 3.9 cm). Mean relative error of LOM-based measurement of true displacement was 20%, maximum — 37%. In some cases, there was an asymmetric movement in the clockwise and counterclockwise directions. It was more noticeable at BBTI 3000 that was associated with the direction of phase-coding gradient.
Our results may be estimated as approximate considering significant variability of data. More extensive experimental base is needed to obtain reliable quantitative results.
Imaging of cerebrospinal fluid flow from the tagging zone in a patient after ventriculostomy is shown in Fig. 2.
Analysis of C1—C2 cervical spine revealed more intense caudal flows. LOM was slightly higher in caudal direction for BBTI 2000 and 3000, but these differences were not significant (Fig. 3b). LOM for BBTI 3000 was 1.64 times higher than for BBTI 2000. These differences were significant in all cases. It should be noted that this difference was smaller than for cerebral aqueduct (1.63 and 2.3 times, respectively). There were significantly higher LOM values for ventral parts compared with dorsal regions in both directions for BBTI 3000 (Fig. 3b).
Our data are generally comparable with the results of Ohtonari T et al. [13]. We found more intense CSF flow in ventral regions compared with dorsal ones. Total LOM (sum of LOM medians for ventral and dorsal surfaces, cranial and caudal directions) was 30.06 mm that is slightly lower than accepted boundary of “normal” values for patients with Chiari-I malformation [14]. The difference may be due various scanning techniques.
We report a 56-year-old patient C. as an example of the use of Time-SLIP MRI for analysis of CSF flow in clinical practice. This patient with occlusive hydrocephalus (adhesion at the level of cerebral aqueduct) underwent ventriculostomy of the 3rd ventricle. Control MRI was performed after 2 months including SSFP (analogue of FIESTA) and Time-SLIP modes (Fig. 4).
Thus, combination of Time-SLIP and SSFP MRI is valuable to analyze CSF circulation after ventriculostomy of the 3rd ventricle and confirm patency of the fistula.
Modification of Time-SLIP MRI was developed. This method does not require cardiac synchronization (scanning time 2:16 min) for analysis of CSF circulation. Mean relative error of LOM-based measurement of true displacement was 20%. This technique should be considered as a screening for rapid diagnosis of impaired cerebrospinal fluid circulation requiring more complex and specialized MRI methods for CSF flow assessment. The technology can also supplement phase-contrast MRI data on impaired cerebrospinal fluid circulation. The developed technique may be demanded for initial assessment of patency of third ventriculostomy, pre- and postoperative analysis of relationships of arachnoid cysts with cerebrospinal fluid spaces. Further large-scale trials are required to confirm clinical significance of our quantitative data.
The authors declare no conflict of interest.
The report of Petryaiykin A.V. et al. is devoted to an urgent problem of non-invasive imaging of cerebrospinal fluid flow in spinal canal and cerebral aqueduct. The authors consider the possibilities of Time-SLIP MRI for imaging of cerebrospinal fluid flow in a phantom and in patients. Time-SLIP MRI is valuable for visual assessment of cerebrospinal fluid flow without cardiac synchronization for 2 minutes. Mean relative error of LOM-based measurement of true displacement was 20%.
Analysis of cerebrospinal fluid flow in cerebral aqueduct and cervical spine of healthy volunteers made it possible to evaluate certain features in these departments (turbulent motion in cerebral aqueduct and different flow of cerebrospinal fluid in caudal and ventral parts of the neck). The only example of the use of Time-SLIP MRI for assessment of cerebrospinal fluid flow after endoscopic third ventriculostomy in a patient with occlusive hydrocephalus allows us to hope for successful clinical implementation of this technique.
However, this approach may be applied as a screening method for rapid diagnosis of impaired cerebrospinal fluid circulation and subsequent more complex and specialized MRI methods for CSF flow assessment. This technology may be especially valuable for initial assessment of patency of third ventriculostomy, pre- and postoperative analysis of relationships of arachnoid cysts with cerebrospinal fluid spaces. Further large-scale trials are required to confirm clinical significance of our quantitative data (LOM).
It would be reasonable to change and concretize the conclusions of the study. Authors’ own opinion regarding the prospects of using this technique in clinical and scientific researches would be desirable.
I.N. Pronin (Moscow, Russia)
References:
Email Confirmation
An email was sent to test@gmail.com with a confirmation link. Follow the link from the letter to complete the registration on the site.
Email Confirmation
We use cооkies to improve the performance of the site. By staying on our site, you agree to the terms of use of cооkies. To view our Privacy and Cookie Policy, please. click here.
