DOI: 10.1148/rg.26si065512
RadioGraphics 2006;26:S97-S110
© RSNA, 2006
Multichannel CT: Evaluating the Spine in Postoperative Patients with Orthopedic Hardware1
Annette C. Douglas-Akinwande, MD,
Kenneth A. Buckwalter, MD,
Jonas Rydberg, MD,
James L. Rankin, RT and
Robert H. Choplin, MD
1 From the Department of Radiology, Indiana University Medical Center, University Hospital 0279, 550 N University Blvd, Indianapolis, IN 46202. Presented as an education exhibit at the 2005 RSNA Annual Meeting. Received April 3, 2006; revision requested June 23 and received July 14; accepted July 24. A.C.D.A., K.A.B., J.R., and R.H.C. supported by educational grants from Philips Medical Systems, Cleveland, Ohio; J.L.R. is a member of the Philips CT Medical Advisory Board, Cleveland, Ohio; and A.C.D.A. is a consultant with Bracco Diagnostics, Princeton, NJ.
Address correspondence to A.C.D.A. (e-mail: andougla{at}iupui.edu).
 |
Abstract
|
|---|
Evaluating the spine in patients with metal orthopedic hardware is challenging. Although the effectiveness of conventional computed tomography (CT) can be limited by severe beam-hardening artifacts, the evolution of multichannel CT in recent years has made available new techniques that can help minimize these artifacts. Multichannel CT allows faster scanning times, resulting in reduced motion artifacts; thinner sections, with which it is possible to create a scanned volume of isotropic voxels with equivalent image resolution in all planes; and the generation of a higher x-ray tube current, which may result in better penetration of metal hardware and reduction of artifacts. Although 140 kVp and high milliamperage-second exposure are recommended for imaging patients with hardware, caution should always be exercised, particularly in children, young adults, and patients undergoing multiple examinations. The acquisition of multiplanar reformatted images in the axial, sagittal, coronal, and oblique planes and of three-dimensional volume-rendered images optimizes image interpretation. Wide window settings are best for reviewing images when hardware is present. The integrity of hardware is best assessed with multiplanar average intensity projection. Soft-tissue structures are best visualized by interactively varying the window width and level settings. Implementation of these techniques can yield diagnostic-quality images and aid in patient treatment.
© RSNA, 2006
|
LEARNING OBJECTIVES FOR TEST 5
|
|---|
After reading this article and taking the test, the reader will be able to:- List the basic principles of multichannel CT.
- Describe the physical basis of metal-related artifacts at CT and design protocols that reduce these artifacts.
- Discuss the clinical relevance of spine imaging in postoperative patients with orthopedic hardware.
|
Introduction
|
|---|
Currently in North America the majority of spine fusion procedures are performed with metal hardware, which increases the fusion rate and provides early stability while the osseous fusion matures (1). Radiography, magnetic resonance (MR) imaging, and computed tomography (CT) can be used to visualize the anatomy of the spine and identify the source of any clinical problems. The projectional nature of radiography limits the amount of information it can provide, whereas cross-sectional imaging with MR imaging and CT can be more helpful. Therefore, MR imaging and CT are the primary imaging modalities used for evaluating the spine (2). However, the presence of metal-related artifacts at both conventional CT (Fig 1) and MR imaging can obscure relevant anatomy and disease. Over the past decade, helical CT has evolved from a single channel to 64 channels (Table 1). This new technology has introduced techniques that can be used to minimize metal-related artifacts. In this article, we review the basic principles of multichannel CT and discuss metal-related artifacts in terms of their physical basis and related imaging considerations. In addition, we discuss and illustrate image acquisition and reformation techniques that can be used to minimize artifacts. We also discuss the clinical relevance of postoperative imaging and the clinical applications of multichannel CT in patients with orthopedic hardware.
View larger version (123K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1a. Metal-related artifacts at conventional CT. (a) CT scan (soft-tissue windowing) shows severe artifacts that obscure anatomy and limit evaluation of the spine and hardware. (b) CT scan (bone windowing) shows reduction of the artifacts; however, the image remains nondiagnostic.
|
|
View larger version (100K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1b. Metal-related artifacts at conventional CT. (a) CT scan (soft-tissue windowing) shows severe artifacts that obscure anatomy and limit evaluation of the spine and hardware. (b) CT scan (bone windowing) shows reduction of the artifacts; however, the image remains nondiagnostic.
|
|
|
Basic Principles of Multichannel CT
|
|---|
Terminology
The capacity of a CT scanner to collect information from multiple contiguous anatomic locations in the z axis during one gantry rotation has been expressed as multislice CT, multisection CT, multidetector CT, multi–detector row CT, and multichannel CT (Table 2). The terms multislice and multisection emphasize, albeit somewhat ambiguously, the number of images that are generated. The terms multidetector and multi–detector row refer to the number of physical detector rows that can register data. However, the number of rows is not a direct indicator of the potential capability of a CT scanner. The term multichannel more accurately refers to the capacity of a scanner to register information about the scanned object during a gantry rotation by means of its data acquisition system (DAS) (3).
Detector Array Configuration
The fundamental component of multichannel CT is the slip-ring design of the detector array and DAS. This design allows the electronic manipulation of the incident x-ray beam into multiple channels of data, whereby simultaneous rows of section information are collected during each gantry rotation (3–5).
Multichannel CT scanners have detector arrays that allow partition of the x-ray beam into multiple combinations of subdivided channels (3,6). This partition is accomplished through the DAS, which electronically controls the circuit that collects and transmits data from the detector rows. For example, the detector array of a 16-channel scanner can have 16, 24, or 40 detector rows; however, the DAS may be used to activate only 16 detector rows (Fig 2a) or all 24 detector rows (Fig 2b) on one vendor’s 16-channel scanner (Brilliance 16; Philips Medical Systems, Cleveland, Ohio). The numbers 2, 4, 8, 10, 16, 32, 40, or 64 when applied to a CT scanner refer to the number of electronic channels (ie, the configuration of the DAS) in the detector package, not necessarily the number of sections generated for each gantry rotation or the number of detector rows. On the basis of the number of available channels, conclusions can be drawn about the functional capability of the CT scanner.
View larger version (57K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2a. Detector array configuration and collimation. (a) Schematic illustrates a 16-channel scanner operating in the 12-mm beam collimation mode. Note that the nominal and reconstructed section thicknesses are small. (b) Schematic illustrates the same scanner operating in the 24-mm beam collimation mode. Note how the smaller central detectors are paired to increase the nominal section thickness. The possible reconstructed sections from raw data are thicker than with the 12-mm beam collimation mode.
|
|
View larger version (61K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2b. Detector array configuration and collimation. (a) Schematic illustrates a 16-channel scanner operating in the 12-mm beam collimation mode. Note that the nominal and reconstructed section thicknesses are small. (b) Schematic illustrates the same scanner operating in the 24-mm beam collimation mode. Note how the smaller central detectors are paired to increase the nominal section thickness. The possible reconstructed sections from raw data are thicker than with the 12-mm beam collimation mode.
|
|
Collimation
Beam Collimation.—
With single-section CT, section thickness is directly related to the physical collimation of the x-ray beam by metal dividers (collimators). Similarly, with multichannel CT, the term beam collimation refers to the work done by metal collimators near the x-ray source. These metal collimators may be adjusted to define the width of the x-ray beam that is directed through the patient. For example, for one particular 16-channel scanner (Brilliance 16, Philips Medical Systems) with two beam collimation modes, the narrow (12-mm) collimation mode uses the 16 smaller central detectors (16 x 0.75 mm) (Fig 2a), whereas the wide (24-mm) collimation mode exposes all 24 detectors (16 x 1.5 mm) (Fig 2b). Beam collimation, along with pitch and gantry rotation speed, influences anatomic coverage.
Section Collimation.—
With multichannel CT, in addition to beam collimation, there is section collimation. Section collimation is the smallest section thickness that can be reconstructed from the acquired data and is based on how the individual detectors are configured to channel the data. With the 16-channel scanner described earlier, when the wide (24-mm) collimation mode (Fig 2b) is used, the smaller (0.75-mm) central detectors are paired, with each pair collecting data as a single 1.5-mm collimator. The nominal section thickness (section collimation) that is possible in this instance is 1.5 mm. Hence, reconstructed sections from such an acquisition cannot be less than 1.5 mm thick. From the data acquired using this mode, sections with a thickness of 2, 3, or 5 mm may be reconstructed. However, when the narrow (12-mm) collimation mode is used (Fig 2a), the smaller central detectors individually collect data; thus, the smallest possible reconstructed thickness is 0.75 mm. From the data acquired using this mode, sections with a thickness of 0.8, 1, 1.5, 2, or 3 mm may be reconstructed.
Effective Section Thickness
With conventional (nonhelical) CT, section width is either a direct function of beam collimation (single-channel CT) or directly related to the width of the detector row (multichannel CT). However, with helical CT, the minimum section thickness is always greater than the width of the collimation (single-channel CT) or the narrowest detector row width (multichannel CT); hence the term effective section thickness. The difference between nominal and effective section thickness is due to section broadening, which results from movement of the scanning table during gantry rotation. The reconstruction algorithm must recreate the exact transverse planes through the patient’s body, but errors can be introduced during the reconstruction of helical data. These errors lead to sections that are slightly thicker than the nominal width. The resulting section thickness is referred to as the "effective" section thickness; sections thicker than the minimum section width may not be precise, depending on the z-axis reconstruction filter supplied by the vendor. 
Thin sections facilitate postprocessing of high-quality multiplanar reformatted (MPR) and volume-rendered (VR) images; therefore, the effective section thickness should be taken into consideration when designing imaging protocols.
Isotropic Imaging
In the context of imaging, the term isotropic simply means that the voxels that make up the volume data set are cubic (Fig 3). This isotropy is achieved when the section thickness approaches the in-plane resolution of the source images. Isotropic imaging optimizes the postprocessing of (for example) MPR, VR, and shaded-surface-display images without loss of spatial resolution.
View larger version (20K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3. Drawings illustrate voxels acquired with single-channel scanners versus 4–64-channel scanners. Voxels acquired with single-channel scanners (left) are anisotropic, and reformatted images have lower resolution than the axial source images. However, 4–64-channel scanners yield isotropic voxels (right), which facilitate reformation with image resolution similar to that of the source images. X, Y, and Z represent the three dimensions of the voxel.
|
|
As a result, multiple optimized views (axial, sagittal, coronal, oblique) may be retrospectively created from a single volume of data obtained from a single acquisition. The ability to obtain MPR images without loss of spatial resolution is a major advance in CT because it minimizes the importance of patient positioning and obviates direct coronal and sagittal imaging, which are not possible in routine imaging of the spine. The optimal plane for viewing disease can always be obtained retrospectively.
Pitch
With single-section helical CT, the term pitch refers to the ratio of table translation (in millimeters) per gantry rotation to the nominal section thickness (7,8). When pitch equals 1, there is neither overlap nor gap in the radiation beam as it moves through the patient; a pitch of less than 1 means that there is overlap in the radiation beam; and a pitch greater than 1 means that there are gaps in the radiation beam. Multichannel CT vendor preferences have led to two definitions of pitch.
According to the first definition, pitch equals table translation (in millimeters) per gantry rotation divided by the width of one detector channel of an N-channel detector array.
However, the nominal section thickness of a multichannel system can be altered by using different channel combinations, possibly resulting in the alteration of pitch for a constant table speed. Therefore, another definition of pitch that can be applied to both single- and multichannel systems is needed.
According to this second definition, pitch equals table translation (in millimeters) per gantry rotation divided by beam collimation. This definition maintains the relationship between pitch and image quality and is the preferred definition in helical scanning (7,8). All CT manufacturers have agreed to and are currently using this second definition.
|
Metal-related Artifacts at CT
|
|---|
Physical Basis
The generation of artifacts at CT from metal hardware is related to the image reconstruction algorithm (filter); tube current (in milliampere-seconds); x-ray kilovolt peak; pitch; and hardware composition, geometry (shape), and location.
Imaging Considerations
CT of orthopedic hardware results in altered x-ray beam attenuation, which creates artifacts on the reconstructed image. If these artifacts are left uncorrected, the resulting images can be nondiagnostic in the vicinity of the hardware. In the past, postprocessing correction techniques centered on filtered back projection, with linear functions used to fill in (9) or avoid (10,11) the missing data. These special image reconstruction techniques are generally unavailable on clinical CT scanners. Filtered back projection is prone to artifacts, which are most severe when the x-ray projections contain missing or noisy data and when there is patient motion. The typical star-shaped artifacts generated during the imaging of patients with metal hardware are partly attributable to the filtered back projection reconstruction method. Metal hardware attenuates the x-ray beam, distorting the projection data and contributing to image artifacts. The choice of reconstruction filter greatly affects the appearance of metal-related artifacts. Some investigators have used edge-enhancement filters (eg, bone algorithm), which accentuates metal-related artifacts (9,11,12), whereas (although it may seem counterintuitive) smooth reconstruction filters (soft-tissue algorithm) greatly reduce these artifacts (Fig 4) (2).
View larger version (146K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4a. Appearance of artifacts with use of an edge-enhancement algorithm versus a smooth algorithm. (a) Sagittal MPR image acquired with an edge-enhancement algorithm from postmyelographic CT data obtained on a 64-channel scanner shows artifacts at L3–L4. (b) Sagittal MPR image acquired with a smooth algorithm minimizes the artifacts. (c) Curved coronal MPR image shows disk herniation, nerve root impingement, and spinal stenosis at L3–L4 (arrow). (d, e) Coronal multiplanar maximum intensity projection (d) and average intensity projection (AIP) (e) images clearly depict intact Harrington rods.
|
|
View larger version (104K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4b. Appearance of artifacts with use of an edge-enhancement algorithm versus a smooth algorithm. (a) Sagittal MPR image acquired with an edge-enhancement algorithm from postmyelographic CT data obtained on a 64-channel scanner shows artifacts at L3–L4. (b) Sagittal MPR image acquired with a smooth algorithm minimizes the artifacts. (c) Curved coronal MPR image shows disk herniation, nerve root impingement, and spinal stenosis at L3–L4 (arrow). (d, e) Coronal multiplanar maximum intensity projection (d) and average intensity projection (AIP) (e) images clearly depict intact Harrington rods.
|
|
View larger version (85K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4c. Appearance of artifacts with use of an edge-enhancement algorithm versus a smooth algorithm. (a) Sagittal MPR image acquired with an edge-enhancement algorithm from postmyelographic CT data obtained on a 64-channel scanner shows artifacts at L3–L4. (b) Sagittal MPR image acquired with a smooth algorithm minimizes the artifacts. (c) Curved coronal MPR image shows disk herniation, nerve root impingement, and spinal stenosis at L3–L4 (arrow). (d, e) Coronal multiplanar maximum intensity projection (d) and average intensity projection (AIP) (e) images clearly depict intact Harrington rods.
|
|
View larger version (51K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4d. Appearance of artifacts with use of an edge-enhancement algorithm versus a smooth algorithm. (a) Sagittal MPR image acquired with an edge-enhancement algorithm from postmyelographic CT data obtained on a 64-channel scanner shows artifacts at L3–L4. (b) Sagittal MPR image acquired with a smooth algorithm minimizes the artifacts. (c) Curved coronal MPR image shows disk herniation, nerve root impingement, and spinal stenosis at L3–L4 (arrow). (d, e) Coronal multiplanar maximum intensity projection (d) and average intensity projection (AIP) (e) images clearly depict intact Harrington rods.
|
|
View larger version (46K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4e. Appearance of artifacts with use of an edge-enhancement algorithm versus a smooth algorithm. (a) Sagittal MPR image acquired with an edge-enhancement algorithm from postmyelographic CT data obtained on a 64-channel scanner shows artifacts at L3–L4. (b) Sagittal MPR image acquired with a smooth algorithm minimizes the artifacts. (c) Curved coronal MPR image shows disk herniation, nerve root impingement, and spinal stenosis at L3–L4 (arrow). (d, e) Coronal multiplanar maximum intensity projection (d) and average intensity projection (AIP) (e) images clearly depict intact Harrington rods.
|
|
Postprocessing algorithms to reduce artifacts (eg, "adaptive filter" [Philips Medical Systems], which recognizes and reduces severe beam-hardening artifacts) may also be applied. Different algorithms from other CT vendors such as "extended windows" (GE Medical Systems, Waukesha, Wis) may also be available; therefore, the reader should become familiar with the capability of the scanners at his or her institution. Postprocessing the raw data with the acquisition of MPR and three-dimensional VR images may reduce metal-related artifacts and allow the assessment of alignment (Fig 5) (12,13). The integrity of hardware can easily be assessed with multiplanar AIP (Fig 6). This postprocessing technique projects the mean attenuation value of the pixels on every view throughout the volume data set onto a two-dimensional image.
View larger version (107K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5a. Reduction of metal-related artifacts and assessment of spinal alignment with postprocessing. (a–c) Axial CT scans obtained at L1 (a), L1–L2 (b), and L3 (c) show extensive bone destruction. Minimal artifact from the orthopedic hardware does not obscure the spinal canal or bone disease. (d) Coronal MPR image shows complete compression of the vertebral body. (e) Sagittal MPR image shows dislocation of the vertebral body. (f) Three-dimensional VR image shows further reduction of the artifacts and provides a general overview of the region of interest, which is helpful for surgical planning. Postprocessing with three-dimensional VR images and coronal and sagittal MPR images is essential for image interpretation.
|
|
View larger version (93K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5b. Reduction of metal-related artifacts and assessment of spinal alignment with postprocessing. (a–c) Axial CT scans obtained at L1 (a), L1–L2 (b), and L3 (c) show extensive bone destruction. Minimal artifact from the orthopedic hardware does not obscure the spinal canal or bone disease. (d) Coronal MPR image shows complete compression of the vertebral body. (e) Sagittal MPR image shows dislocation of the vertebral body. (f) Three-dimensional VR image shows further reduction of the artifacts and provides a general overview of the region of interest, which is helpful for surgical planning. Postprocessing with three-dimensional VR images and coronal and sagittal MPR images is essential for image interpretation.
|
|
View larger version (120K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5c. Reduction of metal-related artifacts and assessment of spinal alignment with postprocessing. (a–c) Axial CT scans obtained at L1 (a), L1–L2 (b), and L3 (c) show extensive bone destruction. Minimal artifact from the orthopedic hardware does not obscure the spinal canal or bone disease. (d) Coronal MPR image shows complete compression of the vertebral body. (e) Sagittal MPR image shows dislocation of the vertebral body. (f) Three-dimensional VR image shows further reduction of the artifacts and provides a general overview of the region of interest, which is helpful for surgical planning. Postprocessing with three-dimensional VR images and coronal and sagittal MPR images is essential for image interpretation.
|
|
View larger version (61K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5d. Reduction of metal-related artifacts and assessment of spinal alignment with postprocessing. (a–c) Axial CT scans obtained at L1 (a), L1–L2 (b), and L3 (c) show extensive bone destruction. Minimal artifact from the orthopedic hardware does not obscure the spinal canal or bone disease. (d) Coronal MPR image shows complete compression of the vertebral body. (e) Sagittal MPR image shows dislocation of the vertebral body. (f) Three-dimensional VR image shows further reduction of the artifacts and provides a general overview of the region of interest, which is helpful for surgical planning. Postprocessing with three-dimensional VR images and coronal and sagittal MPR images is essential for image interpretation.
|
|
View larger version (74K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5e. Reduction of metal-related artifacts and assessment of spinal alignment with postprocessing. (a–c) Axial CT scans obtained at L1 (a), L1–L2 (b), and L3 (c) show extensive bone destruction. Minimal artifact from the orthopedic hardware does not obscure the spinal canal or bone disease. (d) Coronal MPR image shows complete compression of the vertebral body. (e) Sagittal MPR image shows dislocation of the vertebral body. (f) Three-dimensional VR image shows further reduction of the artifacts and provides a general overview of the region of interest, which is helpful for surgical planning. Postprocessing with three-dimensional VR images and coronal and sagittal MPR images is essential for image interpretation.
|
|
View larger version (85K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5f. Reduction of metal-related artifacts and assessment of spinal alignment with postprocessing. (a–c) Axial CT scans obtained at L1 (a), L1–L2 (b), and L3 (c) show extensive bone destruction. Minimal artifact from the orthopedic hardware does not obscure the spinal canal or bone disease. (d) Coronal MPR image shows complete compression of the vertebral body. (e) Sagittal MPR image shows dislocation of the vertebral body. (f) Three-dimensional VR image shows further reduction of the artifacts and provides a general overview of the region of interest, which is helpful for surgical planning. Postprocessing with three-dimensional VR images and coronal and sagittal MPR images is essential for image interpretation.
|
|
View larger version (135K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6a. Assessment of hardware integrity. (a, b) Axial wide-window (a) and sagittal MPR (b) images of the lumbar spine show fixation hardware at L4–L5, but it is difficult to assess whether the metal is intact. (c–e) Axial (c), sagittal (d), and coronal (e) multiplanar AIP images more clearly show the absence of fracture in the hardware. Assessment of hardware integrity is easily achieved with multiplanar AIP.
|
|
View larger version (123K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6b. Assessment of hardware integrity. (a, b) Axial wide-window (a) and sagittal MPR (b) images of the lumbar spine show fixation hardware at L4–L5, but it is difficult to assess whether the metal is intact. (c–e) Axial (c), sagittal (d), and coronal (e) multiplanar AIP images more clearly show the absence of fracture in the hardware. Assessment of hardware integrity is easily achieved with multiplanar AIP.
|
|
View larger version (93K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6c. Assessment of hardware integrity. (a, b) Axial wide-window (a) and sagittal MPR (b) images of the lumbar spine show fixation hardware at L4–L5, but it is difficult to assess whether the metal is intact. (c–e) Axial (c), sagittal (d), and coronal (e) multiplanar AIP images more clearly show the absence of fracture in the hardware. Assessment of hardware integrity is easily achieved with multiplanar AIP.
|
|
View larger version (80K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6d. Assessment of hardware integrity. (a, b) Axial wide-window (a) and sagittal MPR (b) images of the lumbar spine show fixation hardware at L4–L5, but it is difficult to assess whether the metal is intact. (c–e) Axial (c), sagittal (d), and coronal (e) multiplanar AIP images more clearly show the absence of fracture in the hardware. Assessment of hardware integrity is easily achieved with multiplanar AIP.
|
|
View larger version (80K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6e. Assessment of hardware integrity. (a, b) Axial wide-window (a) and sagittal MPR (b) images of the lumbar spine show fixation hardware at L4–L5, but it is difficult to assess whether the metal is intact. (c–e) Axial (c), sagittal (d), and coronal (e) multiplanar AIP images more clearly show the absence of fracture in the hardware. Assessment of hardware integrity is easily achieved with multiplanar AIP.
|
|
|
Multichannel CT Scanning Techniques
|
|---|
The alteration of certain factors in scanning protocols prior to data acquisition affects the ability of the x-ray beam to penetrate metal and reduce metal-related artifacts.
X-ray Kilovolt Peak
A higher x-ray kilovolt peak can increase the likelihood that the x-ray beam will penetrate metal (9). At our institution, we typically use 140 kVp when scanning patients with metal hardware. However, for imaging of the cervical spine in thin patients, 120 kVp may be used to obtain diagnostic-quality images (Fig 7).
View larger version (91K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7a. CT of the cervical spine in a thin patient. Imaging was performed at 120 kVp on a 16-channel scanner. The cross-sectional areas of both the patient and the hardware are small, and the thinnest part of the hardware is perpendicular to the incident x-ray beam. (a, b) Axial (a) and sagittal (b) MPR images (narrow windowing) show prominent metal-related artifacts (arrow in a). (c, d) Axial (c) and sagittal (d) MPR images (wide windowing) minimize the artifacts and are of diagnostic quality. There is a manufactured opening (straight white arrow) in the anterior orthopedic plate and a bone fragment (arrowhead) at C4. An osteophyte at C4–C5 (black arrow) results in spinal canal stenosis (curved white arrow).
|
|
View larger version (73K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7b. CT of the cervical spine in a thin patient. Imaging was performed at 120 kVp on a 16-channel scanner. The cross-sectional areas of both the patient and the hardware are small, and the thinnest part of the hardware is perpendicular to the incident x-ray beam. (a, b) Axial (a) and sagittal (b) MPR images (narrow windowing) show prominent metal-related artifacts (arrow in a). (c, d) Axial (c) and sagittal (d) MPR images (wide windowing) minimize the artifacts and are of diagnostic quality. There is a manufactured opening (straight white arrow) in the anterior orthopedic plate and a bone fragment (arrowhead) at C4. An osteophyte at C4–C5 (black arrow) results in spinal canal stenosis (curved white arrow).
|
|
View larger version (97K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7c. CT of the cervical spine in a thin patient. Imaging was performed at 120 kVp on a 16-channel scanner. The cross-sectional areas of both the patient and the hardware are small, and the thinnest part of the hardware is perpendicular to the incident x-ray beam. (a, b) Axial (a) and sagittal (b) MPR images (narrow windowing) show prominent metal-related artifacts (arrow in a). (c, d) Axial (c) and sagittal (d) MPR images (wide windowing) minimize the artifacts and are of diagnostic quality. There is a manufactured opening (straight white arrow) in the anterior orthopedic plate and a bone fragment (arrowhead) at C4. An osteophyte at C4–C5 (black arrow) results in spinal canal stenosis (curved white arrow).
|
|
View larger version (74K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7d. CT of the cervical spine in a thin patient. Imaging was performed at 120 kVp on a 16-channel scanner. The cross-sectional areas of both the patient and the hardware are small, and the thinnest part of the hardware is perpendicular to the incident x-ray beam. (a, b) Axial (a) and sagittal (b) MPR images (narrow windowing) show prominent metal-related artifacts (arrow in a). (c, d) Axial (c) and sagittal (d) MPR images (wide windowing) minimize the artifacts and are of diagnostic quality. There is a manufactured opening (straight white arrow) in the anterior orthopedic plate and a bone fragment (arrowhead) at C4. An osteophyte at C4–C5 (black arrow) results in spinal canal stenosis (curved white arrow).
|
|
X-ray Tube Current
A higher tube current setting may increase the likelihood that the x-ray beam will penetrate metal, and these settings are possible with multichannel CT because of the ability to use lower pitch settings. The CT x-ray tube contains both a large and a small focal spot. Choosing the large focal spot allows use of the highest exposure factors, which may also improve penetration of the metal hardware by the x rays. However, the compromise is reduced spatial resolution.
Pitch
With multichannel CT, a lower pitch setting allows the collection of redundant data, thereby increasing the likelihood that adequate projection data will be collected. In addition, lower pitch settings reduce splay artifacts—radially alternating bands of higher- and lower-attenuation projection across the image—which are inherent in multichannel CT. Splay artifacts appear as rotating windmill or propeller blades with sharp edges and are particularly conspicuous when one is scrolling through an image stack. These artifacts are introduced by the z-axis sampling errors inherent in helical CT. The artifacts can be eliminated if the reconstructed section width is at least twice the detector element width; however, this will result in z-axis blurring and degradation of MPR images. Splay artifacts can also be reduced by using very low pitch settings (eg, 0.3). Given that scanner setup and technical factors remain unchanged, splay artifacts are reduced as the number of detector rows increases. Thus, all other factors being equal, a 16-channel scanner will produce more splay artifacts than a 64-channel scanner.
Metal Composition
The artifacts generated by orthopedic hardware are related to the composition of the hardware. Materials with lower x-ray beam attenuation coefficients (density) produce fewer artifacts (eg, plastic < titanium < vitallium < stainless steel < cobalt-chrome) (2,11,14–16). Thus, artifacts from titanium, a less dense metal, are minimal, whereas some of the most severe artifacts are produced by cobalt-chrome. Knowledge of the composition of the orthopedic hardware prior to CT is important because the technical parameters of the imaging protocol may be adjusted to minimize radiation exposure to patients. If this information is unavailable, it is possible to review the CT scout images to assess the overall density of the hardware, which is proportional to its attenuation.
Geometric Factors
The amount of artifact generated at CT is also related to the cross-sectional area of the hardware. Orthopedic hardware with a smaller cross-sectional area results in lower x-ray beam attenuation, thus producing fewer missing projection data and fewer artifacts. If possible, the affected body part should be positioned so that the x-ray beam traverses the metal at its smallest cross-sectional area—although the repositioning of body parts is generally not possible for spine imaging. X-ray beam attenuation is also greatest in the regions of greatest patient girth and bone mass. Hence, x-ray beam attenuation in the cervical spine is less than that in the thoracic spine, which in turn is less than that in the lumbar and sacral spine.
The patient’s arms should be positioned to avoid traversal by the x-ray beam, thereby reducing tissue attenuation. At our institution, a traction device is used to lower the patient’s shoulders for imaging of the cervical spine (Fig 8). For imaging of the thoracic, lumbar, and sacral spine, the arms are raised above the head.
View larger version (145K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8. Photograph shows a patient positioned in the CT scanner for cervical spine imaging. The chin is hyperextended, and a shoulder traction device (arrows) is attached to the wrist and wrapped under the soles of the feet. Extension of the knees lowers the shoulders with slow steady traction, helping to minimize beam-hardening artifacts from the shoulders.
|
|
|
Clinical Relevance of Postoperative Imaging
|
|---|
Postoperative imaging of the spine is performed for (a) routine assessment of patients with a good clinical outcome or (b) assessment of recurrent or intractable pain (1).
Routine Assessment
After spinal fusion surgery, imaging is routinely performed to evaluate for the development of pseudoarthrosis or to determine the degree of spinal fusion. This information is used as a guide for the initiation of physical therapy and the elimination of restrictive bracing (1).
Assessment of Recurrent or Intractable Pain
After spine surgery, patients may have minimal or no relief of symptoms and may develop new or recurrent symptoms either shortly or long after initial surgery. Failed back surgery syndrome is a general term used with reference to patients with recurrent symptoms or in whom initial surgery failed to correct the problem (17). This syndrome is seen in 10%–40% of patients who undergo back surgery (18). However, failed back surgery syndrome is a clinical syndrome that does not address the cause of the symptoms. Affected patients often require cross-sectional imaging to determine the cause of the problem and the next therapeutic intervention. The cause may be any combination of the following entities: incorrect placement of pedicle screws (19), fracture of metal hardware, posterior displacement of bone into the spinal canal (Fig 9) (18), failure of fusion with development of a pseudoarthrosis (Fig 9), spondylolysis, osteophytosis, arachnoiditis, recurrent disk herniation, fibrosis (Fig 10), spinal or lateral recess stenosis, infection, or diskitis (1,18). Although disk herniation is typically hyperattenuating (90–120 HU) and fibrosis usually has lower attenuation (50–75 HU), occasionally there may be some overlap, which can make visual distinction between the two conditions difficult (18).
View larger version (113K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9a. Multilevel fusion of the cervical spine. (a, b) CT scans obtained with soft-tissue (a) and bone (b) windowing on a 64-channel scanner at the level of screws within a vertebral body allow excellent visualization of the spinal canal without superimposed artifact. (c–e) Coronal MPR image (c) and sagittal MPR images obtained with soft-tissue (d) and bone (e) windowing show posterior subluxation of C3 on C4 (arrow in d and e), osteopenic bones, odontoid erosions, and compressed vertebral bodies. Multiple manufactured openings are seen in the anterior plate (arrowheads in e).
|
|
View larger version (102K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9b. Multilevel fusion of the cervical spine. (a, b) CT scans obtained with soft-tissue (a) and bone (b) windowing on a 64-channel scanner at the level of screws within a vertebral body allow excellent visualization of the spinal canal without superimposed artifact. (c–e) Coronal MPR image (c) and sagittal MPR images obtained with soft-tissue (d) and bone (e) windowing show posterior subluxation of C3 on C4 (arrow in d and e), osteopenic bones, odontoid erosions, and compressed vertebral bodies. Multiple manufactured openings are seen in the anterior plate (arrowheads in e).
|
|
View larger version (65K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9c. Multilevel fusion of the cervical spine. (a, b) CT scans obtained with soft-tissue (a) and bone (b) windowing on a 64-channel scanner at the level of screws within a vertebral body allow excellent visualization of the spinal canal without superimposed artifact. (c–e) Coronal MPR image (c) and sagittal MPR images obtained with soft-tissue (d) and bone (e) windowing show posterior subluxation of C3 on C4 (arrow in d and e), osteopenic bones, odontoid erosions, and compressed vertebral bodies. Multiple manufactured openings are seen in the anterior plate (arrowheads in e).
|
|
View larger version (94K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9d. Multilevel fusion of the cervical spine. (a, b) CT scans obtained with soft-tissue (a) and bone (b) windowing on a 64-channel scanner at the level of screws within a vertebral body allow excellent visualization of the spinal canal without superimposed artifact. (c–e) Coronal MPR image (c) and sagittal MPR images obtained with soft-tissue (d) and bone (e) windowing show posterior subluxation of C3 on C4 (arrow in d and e), osteopenic bones, odontoid erosions, and compressed vertebral bodies. Multiple manufactured openings are seen in the anterior plate (arrowheads in e).
|
|
View larger version (83K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9e. Multilevel fusion of the cervical spine. (a, b) CT scans obtained with soft-tissue (a) and bone (b) windowing on a 64-channel scanner at the level of screws within a vertebral body allow excellent visualization of the spinal canal without superimposed artifact. (c–e) Coronal MPR image (c) and sagittal MPR images obtained with soft-tissue (d) and bone (e) windowing show posterior subluxation of C3 on C4 (arrow in d and e), osteopenic bones, odontoid erosions, and compressed vertebral bodies. Multiple manufactured openings are seen in the anterior plate (arrowheads in e).
|
|
View larger version (105K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10a. Fibrosis. (a, b) Postmyelographic CT scans obtained with bone (a) and soft-tissue (b) windowing on a four-channel scanner show beam-hardening artifacts from orthopedic hardware. However, the spinal canal and neural foramina are well visualized. (c–e) Sagittal MPR images obtained with soft-tissue (c, d) and bone (e) windowing show a ventrolateral epidural defect at L5-S1 (arrows in c and d), a finding that may represent disk herniation or, as in this case, fibrosis.
|
|
View larger version (124K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10b. Fibrosis. (a, b) Postmyelographic CT scans obtained with bone (a) and soft-tissue (b) windowing on a four-channel scanner show beam-hardening artifacts from orthopedic hardware. However, the spinal canal and neural foramina are well visualized. (c–e) Sagittal MPR images obtained with soft-tissue (c, d) and bone (e) windowing show a ventrolateral epidural defect at L5-S1 (arrows in c and d), a finding that may represent disk herniation or, as in this case, fibrosis.
|
|
View larger version (81K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10c. Fibrosis. (a, b) Postmyelographic CT scans obtained with bone (a) and soft-tissue (b) windowing on a four-channel scanner show beam-hardening artifacts from orthopedic hardware. However, the spinal canal and neural foramina are well visualized. (c–e) Sagittal MPR images obtained with soft-tissue (c, d) and bone (e) windowing show a ventrolateral epidural defect at L5-S1 (arrows in c and d), a finding that may represent disk herniation or, as in this case, fibrosis.
|
|
View larger version (76K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10d. Fibrosis. (a, b) Postmyelographic CT scans obtained with bone (a) and soft-tissue (b) windowing on a four-channel scanner show beam-hardening artifacts from orthopedic hardware. However, the spinal canal and neural foramina are well visualized. (c–e) Sagittal MPR images obtained with soft-tissue (c, d) and bone (e) windowing show a ventrolateral epidural defect at L5-S1 (arrows in c and d), a finding that may represent disk herniation or, as in this case, fibrosis.
|
|
View larger version (66K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10e. Fibrosis. (a, b) Postmyelographic CT scans obtained with bone (a) and soft-tissue (b) windowing on a four-channel scanner show beam-hardening artifacts from orthopedic hardware. However, the spinal canal and neural foramina are well visualized. (c–e) Sagittal MPR images obtained with soft-tissue (c, d) and bone (e) windowing show a ventrolateral epidural defect at L5-S1 (arrows in c and d), a finding that may represent disk herniation or, as in this case, fibrosis.
|
|
|
Clinical Applications of Multichannel CT
|
|---|
Imaging of the postoperative spine requires an understanding of multichannel CT. The protocols typically used at our institution are presented in Tables 3–5; however, the reader may modify these protocols based on the type of scanners available at his or her institution. Optimal scanning parameters depend on the clinical questions to be answered. We recommend using a smooth reconstruction algorithm (soft-tissue filter) to acquire the helical CT data set, since doing so not only minimizes artifacts but also improves soft-tissue evaluation. If there is a concern for infection, contrast material may be administered intravenously.
A kilovolt peak of 140 is recommended for the imaging of spines with orthopedic hardware. Exceptions are made for the imaging of cervical spines and in patients with small cross-sectional areas, in which cases 120 kVp may be used.
In general, imaging of smaller body parts such as the cervical spine benefits from use of the narrowest section thickness. A small section thickness with 50% overlap (eg, 1-mm thickness, 0.5-mm overlap) is best for imaging the cervical spine.
Although the specific value for milliampere-seconds depends on the scanner, the resolution, the hardware geometry, and the patient, some typical values are listed in Tables 3–5. When coverage of a larger anatomic area (eg, thoracic spine or entire spine) is required and 4–16-channel scanners are used, the wide collimation mode is chosen. This compromise is not necessary when 40- or 64-channel scanners are used. Higher milliampere-second exposures and standard resolution are recommended for the imaging of larger patients or of the lumbar spine.
|
Conclusions
|
|---|
Top
Abstract
LEARNING OBJECTIVES FOR TEST...
