Surg Radiol Anat (2010) 32:581–585 DOI 10.1007/s00276-009-0601-0
O R I G I N A L A R T I CL E
Comparison of spinal anatomy between 3-Tesla MRI and CT-myelography under healthy and pathological conditions Astrid Ellen Grams · Jens Gempt · Annette Förschler
Received: 11 October 2009 / Accepted: 16 November 2009 / Published online: 6 December 2009 © Springer-Verlag 2009
Abstract Purpose In many centres both MRI and CT-myelography are performed for treatment planning of degenerative spine disease. More and more centres acquire 3-Tesla MRI scanners in which some artefacts, which lead to diYculties in image evaluation, are more pronounced than at 1.5 Tesla. Aim of this study was to compare spinal physiological and pathological anatomy between 3-Tesla MRI and CT-myelography and to review current imaging standards. Methods In 47 spinal segments commonly used 3-Tesla T2-weighted sequences and CT-myelography studies were evaluated retrospectively. Spinal canal, neural foraminal, spinal cord and disc protrusion diameters were measured. Results The spinal canal was found to be 10% tighter with the utilized MRI sequences, in comparison to CT-M and foraminal diameters were found to be 19.7% tighter in MRI. This was more pronounced in narrowed than in healthy segments. Spinal cord size and size of disc protrusions displayed no signiWcant diVerence between MRI and CT-myelography. Conclusions The main advantage of CT-myelography, in comparison to 3-Tesla MRI, is the reliable information about the bony structures. Soft tissues like the spinal cord A. E. Grams (&) Department of Diagnostic and Interventional Radiology and Neuroradiology, University Hospital Essen, Hufelandstr. 55, 45122 Essen, Germany e-mail:
[email protected] J. Gempt Department of Neurosurgery, Klinikum rechts der Isar, Technical University of Munich, Munich, Germany A. Förschler Department of Neuroradiology, Klinikum Rechts der Isar, Technical University of Munich, Munich, Germany
or disc protrusions were visualised equivalently with both modalities concerning diameters. Keywords 3 Tesla · CT · Myelography · Spinal stenosis · Foraminal stenosis
Introduction In spinal imaging degenerative changes such as bony stenosis or disc protrusions are frequently observed even in asymptomatic patients [12]. Therefore, it remains a challenge to establish diagnosis of clinically relevant nerve root compression or spinal canal stenosis and to enable correct surgical planning, especially in patients who suVer from multisegmental or multicausal degenerative changes or patients who have undergone previous spinal surgery [5]. With computed tomography (CT) imaging bony structures can be visualised reliably; in this regard it is commonly accepted as the “gold standard”. CT imaging after intrathecal injection of an iodine contrast agent, known as CT-myelography (CT-M), provides additional information concerning the eVects of bony structures or soft tissues on the dural sac and the proximal spinal nerve roots [6]. However, CT-M is an invasive examination which comprehends several possible adverse events, such as meningitis, CSF leaks or nerve root injury. Magnetic resonance imaging (MRI) is believed as the “gold standard” in the visualisation of soft tissues, such as disc protrusions, hypertrophy of the ligamentum Xavum, or scar tissue, which may display reasons for spinal or foraminal stenosis. However, the spatial relations of bony structures are often overestimated in MRI [2, 8]. In recent years many centres introduced 3-Tesla scanners for use in clinical routine instead if 1.5-Tesla scanners.
123
582
In general, 3-Tesla imaging is known to provide a higher signal to noise ratio, a better spatial resolution and a shorter acquisition time with a comparable image quality to 1.5-Tesla scanners [7]. On the other hand, at 3 Tesla both susceptibility artefacts and pulsation artefacts, which play an important role in imaging of the spine, are more pronounced compared to 1.5-Tesla imaging [13]. Due to this diVerent pros and cons of the described methods, in some centres it is common to perform both MRI and CT-M to obtain complete imaging information for treatment decisions. Aim of this study was to receive better knowledge of normal and pathological 3-Tesla anatomy of the spine in comparison to CT with intrathecal contrasting, as well as to review the diagnostic value of both methods for the evaluation of bony spinal and neuroforaminal stenosis, as well as disc protrusions.
Patients and methods Imaging studies of 17 patients (9 male and 8 female) who had undergone both CT-M and 3-Tesla T2-weighted MRI imaging of the same spinal segments were evaluated retrospectively. Ten cervical and seven lumbar examinations were compared equivalently. The mean age of the patients was 63 years (range 35–78 years). CT-M and MRI examinations did not exceed more than 3 months. Patients who had spinal surgery in between were excluded. Most patients received the examinations for surgical planning of decompression of a spinal stenosis. Two patients who suVered from neck pain did not show pathological imaging Wndings. For CT-M 10 ml (for lumbar imaging) or 15 ml (for cervical imaging) iopamidol (Solutrast 250 M®, Bracco Altana Pharma, Konstanz) was injected intrathecally via lumbar puncture. Images were achieved using a 64-channel CT scanner (Brilliance 64, Philips, Netherlands, B.V.); 1-mm overcontiguous slices (FOV 250 mm, matrix: 512 £ 512) were reconstructed from a spiral measurement (kv/mAs, 140/250; collimation, 64 £ 0.625; pitch, 0.393) parallel to the disc spaces. Fig. 1 Spinal canal measurement with a axial CT (blue) and b axial MRI (yellow) images of the lumbar spine; c CT and MRI overlay with equal opacities, for comparative visualisation (colour Wgure online)
123
Surg Radiol Anat (2010) 32:581–585
For 3-Tesla MRI, sagittal and axial T2 images used in daily routine were acquired using a 3-Tesla whole-body MR-scanner (Achieva 3 Tesla, Philips, Netherlands, BV) and a RF spine coil. For lumbar spine imaging a T2 turbo spin echo sequence (TR/TE 5465/120, resolution 0.5 £ 0.7 £ 3 mm, slice gap 0.3 mm) was performed; cervical imaging consisted of a 3-D fast Weld echo sequence (TR/TE 17/4 ms; resolution, 0.5 £ 0.7 £ 3 mm without slice gap). Both sequences were acquired parallel to the disc spaces, in addition sagittal sequences were acquired for the easier segmental orientation. As the contrast of signal intensity between CSF and bone is necessary for the reliable determination of the bony borders, it is provided in T2-weighted sequences, T1-weighted sequences were not applied [11]. Forty-seven spinal segments and 94 neural foramina were evaluated. Twelve of the patients did not undergo spinal surgery before the present imaging. Five patients had one or more previous surgical decompressions of the spinal canal. Among the investigated spinal segments 29 stenosed and 20 non-stenosed spinal segments, 33 foraminal stenoses and 61 non-stenosed foramina were found. Twelve segments displayed spinal soft disc protrusions. The classiWcation of “stenosed” versus “non-stenosed” and “disc protrusion” followed the evaluations of two experienced neuroradiologists and an experienced neurosurgeon. In discrepant opinions consensus was found. Diameters were measured in axial planes with the software iplan spine (Brainlab, Heimstetten, Germany), with which a comparative colour-coded visualisation of both applied modalities was possible (Figs. 1, 2). Spinal canal diameters were evaluated at the level of the disc spaces in midline from the dorsal part of the Wbrous annulus to the ventral part of the vertebral lamina. Foraminal diameters were measured at the middle part of the foramen between the bony borders in each segment. Spinal cord diameters were sized at the level of the discs spaces and disc protrusions were measured in anterior–posterior dimension at the most pronounced site. For statistical analysis mean and standard deviations were calculated and paired t tests performed (StatView,
Surg Radiol Anat (2010) 32:581–585
583
Fig. 2 Neural foraminal measurement with a axial CT (blue) and b MRI (yellow) images of the cervical spine; c CT and MRI overlay with equal opacities, for comparative visualisation (colour Wgure online)
Table 1 Spinal canal diameters and discrepancies between MRI and CT-M
n
MRI diameter [mm] (§SD)
CT-M diameter [mm] (§SD)
Discrepancy (mm)
Discrepancy (%)
p value
28
9.03 (§2.5)
10.04 (§2.56)
1.0
10
