Evaluation of monoenergetic imaging to reduce metallic instrumentation artifacts in computed tomography of the cervical spine

Full access

OBJECT

Monoenergetic imaging with dual-energy CT has been proposed to reduce metallic artifacts in comparison with conventional polychromatic CT. The purpose of this study is to systematically evaluate and define the optimal dual-energy CT imaging parameters for specific cervical spinal implant alloy compositions.

METHODS

Spinal fixation rods of cobalt-chromium or titanium alloy inserted into the cervical spine section of an Alderson Rando anthropomorphic phantom were imaged ex vivo with fast-kilovoltage switching CT at 80 and 140 peak kV. The collimation width and field of view were varied between 20 and 40 mm and medium to large, respectively. Extrapolated monoenergetic images were generated at 70, 90, 110, and 130 kiloelectron volts (keV). The standard deviation of voxel intensities along a circular line profile around the spine was used as an index of the magnitude of metallic artifact.

RESULTS

The metallic artifact was more conspicuous around the fixation rods made of cobalt-chromium than those of titanium alloy. The magnitude of metallic artifact seen with titanium fixation rods was minimized at monoenergies of 90 keV and higher, using a collimation width of 20 mm and large field of view. The magnitude of metallic artifact with cobalt-chromium fixation rods was minimized at monoenergies of 110 keV and higher; collimation width or field of view had no effect.

CONCLUSIONS

Optimization of acquisition settings used with monoenergetic CT studies might yield reduced metallic artifacts.

ABBREVIATIONSFOV = field of view; GSI = Gemstone Spectral Imaging; keV = kiloelectron volt; kVp = peak kilovoltage.

OBJECT

Monoenergetic imaging with dual-energy CT has been proposed to reduce metallic artifacts in comparison with conventional polychromatic CT. The purpose of this study is to systematically evaluate and define the optimal dual-energy CT imaging parameters for specific cervical spinal implant alloy compositions.

METHODS

Spinal fixation rods of cobalt-chromium or titanium alloy inserted into the cervical spine section of an Alderson Rando anthropomorphic phantom were imaged ex vivo with fast-kilovoltage switching CT at 80 and 140 peak kV. The collimation width and field of view were varied between 20 and 40 mm and medium to large, respectively. Extrapolated monoenergetic images were generated at 70, 90, 110, and 130 kiloelectron volts (keV). The standard deviation of voxel intensities along a circular line profile around the spine was used as an index of the magnitude of metallic artifact.

RESULTS

The metallic artifact was more conspicuous around the fixation rods made of cobalt-chromium than those of titanium alloy. The magnitude of metallic artifact seen with titanium fixation rods was minimized at monoenergies of 90 keV and higher, using a collimation width of 20 mm and large field of view. The magnitude of metallic artifact with cobalt-chromium fixation rods was minimized at monoenergies of 110 keV and higher; collimation width or field of view had no effect.

CONCLUSIONS

Optimization of acquisition settings used with monoenergetic CT studies might yield reduced metallic artifacts.

ABBREVIATIONSFOV = field of view; GSI = Gemstone Spectral Imaging; keV = kiloelectron volt; kVp = peak kilovoltage.

Since the introduction of spinal fusion in 1911 for the treatment of tuberculosis,1,8 it has commonly been used for the treatment of a wide range of spinal conditions, including deformity, traumatic instability, and a variety of degenerative conditions. Spinal implants—including screws and rods—are manufactured using stainless steel, pure titanium, titanium alloy, and cobalt-chromium alloy. This instrumentation interferes with the quality of CT imaging because of beam hardening (due to the preferential absorption of soft x-rays in the polyenergetic x-ray beam spectrum as x-rays pass through the body), photon starvation (a manifestation of excessive noise in the raw data profile and the nonlinear nature of logarithmic conversion), and scatter artifacts (caused by their high x-ray attenuation coefficient).9,14 Titanium and its metal alloys generate less artifact compared with stainless steel or cobalt-chromium.10 Titanium is more sensitive to notching and rod failure, is less stiff, and is more difficult to bend precisely compared with stainless steel or cobalt-chromium.

The beam hardening, photon starvation, and scatter artifact produced by spinal instrumentation needs to be minimized to assess various components of the spine, including the central canal, neural foramen, disc spaces, vertebral bodies, and paraspinal tissues. Various approaches may diminish the production of artifacts on conventional multidetector CT: use of a high peak voltage, high tube current, narrow collimation, and optimized image reconstruction;11 however, despite advances in detector technology and computer software, artifacts from metal implants remain a problem. Dual-energy CT, allowing for analysis of energy-dependent changes in the attenuation of different materials, has been proposed as a means to reduce beam-hardening metal artifacts via generating monoenergetic images.3,7 This is particularly promising in light of recent studies demonstrating that dual-energy CT allows a reduction of radiation dose with comparable signal-to-noise ratio relative to conventional single-energy acquisitions.4,15 Although dual-energy CT scanners are being increasingly used for imaging of patients with metallic spinal instrumentation, there is little information available regarding the effect of acquisition parameters and hardware material on the severity of artifacts.5,6,13,16

The purpose of the present study was to systematically evaluate and define the optimal dual-energy CT imaging parameters for specific cervical spinal implant alloy compositions.

Methods

Phantom Preparation

Spinal fixation rods of cobalt-chromium or titanium alloys (5.5-mm diameter; Medtronic Sofamor Danek) were inserted into the water-filled cervical spinal canal channel of an Alderson Rando anthropomorphic phantom (Fig. 1). The phantom has been used for more than 30 years in radiology and radiotherapy2 and represents a reference person that was constructed with a natural human skeleton adjusted to overcome any natural lack of symmetry, with distorted joints cast inside a proprietary homogeneous urethane formulation that has attenuation characteristics similar to those of water.

FIG. 1.
FIG. 1.

Photograph of the anthropomorphic phantom in the gantry of the CT scanner. The metallic rod was inserted into the spinal access channel of the phantom (arrow), which was then filled with water and taped. Figure is available in color online only.

Dual-Energy CT Imaging

All images in this study were obtained using Gemstone Spectral Imaging (GSI) technology on a 64-row single-source dual-energy fast-kilovoltage switching CT scanner (Discovery CT750 HD, GE Healthcare) with the following parameters: helical mode, axial plane, 0.625-mm slice thickness, 0.625-mm interval, and dual-energy tube voltages at 80 and 140 kVp (peak kilovoltage). The quality control images (140 kVp) were used as polyenergetic controls. The beam collimation varied between 20 and 40 mm, and the scanning field of view (FOV) was medium and large body (50 cm) and medium head (32 cm), corresponding to presets 10, 11, 12, 19, and 47 (Table 1) available on the scanner. The acquired CT data were transferred to a workstation for postprocessing (AW VolumeShare 5, GE Healthcare).

TABLE 1

Radiation dose and acquisition parameters of dual-energy CT protocols

GSI Preset*FOV Setting (mm)Collimation (mm)PitchRotation Time (sec)Calibration PhantomTube Current in mACTDIvol in mGy
19Medium head (320)200.9690.6Head 1664053.84
47Medium head (320)200.9690.7Head 1627526.41
11Medium body (500)401.3750.8Body 3260018.81
12Large body (500)201.3750.8Body 3260019.63
10Large body (500)401.3750.8Body 3260018.01
CTDIvol = weighted volume CT dose index.

GSI protocol preset names are proprietary manufacturer settings for GE Healthcare and can be selected by the technologist at the time of scanning.

Due to the fast-kilovoltage switching configuration, the provided tube current values are approximate.

Image Analysis

The standard deviation of voxel intensities along a circular line profile around the cervical spine, as depicted in the inset in Fig. 2, was determined using ImageJ (National Institutes of Health).

FIG. 2.
FIG. 2.

Representative montage of axial and sagittal dual-energy CT images of the cervical spine with titanium (Ti) or cobalt-chromium (Co-Cr) alloy fixation rod inserted in the spinal canal of the anthropomorphic phantom with settings indicated on the right (medium or large FOV, 20 and 40-mm collimation), reconstructed at polychromatic (Poly) and calculated monoenergetic values of 70, 90, 110, and 130 keV. Lookup table window center and width were 40 and 350 HU, respectively.

Results

The artifacts radially projecting from the metallic rod inserted into the spinal canal were significantly less conspicuous with titanium alloy than with cobalt-chromium alloy (p < 0.05) (Table 2 and Fig. 2). The artifact was reduced on the extrapolated monoenergetic images of increasing monoenergetic x-ray energies by up to 49% and 54% of that of polychromatic images, for titanium and cobalt-chromium alloys, respectively (Table 2 and Fig. 3).

TABLE 2

Variability (± SD) of voxel intensities along a circular line profile around the spine with metallic instrumentation of either titanium or cobalt-chromium alloy*

X-Ray Energy (keV)Setting
Medium/20/19Medium/20/47Medium/40/11Large/20/12Large/40/10
Titanium
 Polychromatic11.0 ± 2.311.7 ± 2.911.2 ± 2.69.8 ± 2.213.8 ± 6.0
 707.0 ± 0.89.0 ± 1.28.3 ± 1.56.5 ± 1.39.4 ± 1.0
 906.1 ± 0.48.6 ± 0.56.5 ± 0.85.4 ± 0.87.7 ± 0.7
 1105.8 ± 0.38.5 ± 0.75.3 ± 0.64.4 ± 0.76.7 ± 0.8
 1305.7 ± 0.28.6 ± 0.64.7 ± 0.54.1 ± 0.66.4 ± 1.1
Cobalt-chromium
 Polychromatic18.3 ± 5.819.8 ± 4.718.9 ± 5.017.9 ± 4.319.1 ± 5.1
 7019.5 ± 6.516.5 ± 3.721.2 ± 6.721.9 ± 6.822.3 ± 6.6
 9011.9 ± 3.313.7 ± 2.913.0 ± 3.514.0 ± 3.714.2 ± 3.4
 1108.6 ± 1.411.1 ± 1.09.0 ± 1.710.0 ± 1.710.3 ± 1.4
 1307.7 ± 0.810.5 ± 1.07.6 ± 1.08.4 ± 1.08.9 ± 0.8

Data are given as mean ± SD in Hounsfield units. The respective values for titanium alloy were lower than those of cobalt-chromium alloy (p < 0.05, paired t-test).

The settings are as follows: medium or large FOV/20- or 40-mm collimation/GSI presets (proprietary manufacturer settings for GE Healthcare).

FIG. 3.
FIG. 3.

A: Representative graph of voxel intensities along the circular line profile around the spine (see circle in inset). B: Graph depicting the variability of voxel intensities (mean ± SD), an index of artifact conspicuity, as a function of reconstructed x-ray energy, FOV, and collimation when either titanium or cobalt-chromium alloy fixation rods were placed in the spinal canal.

The magnitude of metallic artifact seen with titanium fixation rods was minimized at monoenergies of 90 kiloelectron volts (keV) and higher, using a collimation width of 20 mm and large field of view (a 45%, 55%, and 59% reduction compared with the 80-kVp polychromatic images at 90, 110, and 130 keV, respectively) (Table 2 and Figs. 2 and 3). The magnitude of metallic artifact seen with cobalt-chromium fixation rods was minimized at monoenergies of 110 keV and higher (51% reduction as compared with the 80-kVp polychromatic images); collimation width or field of view had no effect (Table 2 and Figs. 2 and 3).

Discussion

Computed tomography is commonly performed in the imaging evaluation of the spine after surgery, giving excellent bony detail, accurate assessment of both spinal and construct alignment and of instrumentation position relative to the spinal canal and nerve roots, and degree of osseous fusion. The present study demonstrated that reconstruction of images at higher extrapolated monoenergetic x-ray energies significantly diminishes (without completely eliminating) the artifacts radially projecting from the hardware. This is consistent with prior studies3,6,12,18 and likely reflects that the artifacts for very dense metal implants are caused by factors in addition to beam hardening, which cannot be corrected by synthesizing high-energy monoenergetic images. This notion is further supported by the observation that artifact reduction on extrapolated monoenergetic x-ray energies was better seen with instrumentation manufactured from the lower x-ray attenuation titanium alloy compared with higher x-ray attenuation cobalt-chromium alloy. However, we speculate that the fractional reduction in variability at optimal settings versus controls in Fig. 3 underestimates the expected improvement in perceived image quality since the variability represents a combination of residual artifact and noise.

This study was limited by an ex vivo experimental approach and will serve as a guide for additional in vivo studies, which are required to balance the optimal imaging parameters for best osseous and soft-tissue definition in vivo. Since metal and bone (given their high x-ray attenuation coefficient) are responsible for the majority of reconstruction artifacts in CT, bone-metal phantoms are useful in devising techniques to minimize the artifacts; moreover, the homogeneous soft tissue–equivalent material allows for straightforward artifact quantification. Despite the ex vivo design, however, the doses used are well within and average on the spectrum of anticipated doses to patients for these types of examinations.17 Moreover, the novel parameter (to the best of our knowledge, not reported earlier) used to evaluate the severity of artifact (standard deviation of a circular line profile) is potentially confounded by varied amounts of quantum noise in the different images. This can be minimized by smoothing or by filtering out the frequency components corresponding to quantum noise, but these techniques were not used in the present study, given similar quantum mottle in the images subjected to qualitative analysis.

In summary, optimization of acquisition settings used with monoenergetic CT studies might yield reduced metallic artifacts.

Author ContributionsConception and design: all authors. Acquisition of data: all authors. Analysis and interpretation of data: Wintermark, Komlosi, Drafting the article: Wintermark, Komlosi. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Wintermark. Statistical analysis: all authors. Administrative/technical/material support: Wintermark, Komlosi. Study supervision: Wintermark.

References

  • 1

    Albee FH: Transplantation of a portion of the tibia into the spine for Pott's disease: a preliminary report 1911. Clin Orthop Relat Res 460:14162007

    • Search Google Scholar
    • Export Citation
  • 2

    Archer BRWhitmore RCNorth LBBushong SC: Bone marrow dose in chest radiography: the posteroanterior vs. anteroposterior projection. Radiology 133:2112161979

    • Search Google Scholar
    • Export Citation
  • 3

    Bamberg FDierks ANikolaou KReiser MFBecker CRJohnson TR: Metal artifact reduction by dual energy computed tomography using monoenergetic extrapolation. Eur Radiol 21:142414292011

    • Search Google Scholar
    • Export Citation
  • 4

    Bauer RWKramer SRenker MSchell BLarson MCBeeres M: Dose and image quality at CT pulmonary angiography-comparison of first and second generation dualenergy CT and 64-slice CT. Eur Radiol 21:213921472011

    • Search Google Scholar
    • Export Citation
  • 5

    Grams AESender JMoritz RObert MStein MOertel M: Dual energy CT myelography after lumbar osteosynthesis. Rofo 186:6706742014

    • Search Google Scholar
    • Export Citation
  • 6

    Guggenberger RWinklhofer SOsterhoff GWanner GAFortunati MAndreisek G: Metallic artefact reduction with monoenergetic dual-energy CT: systematic ex vivo evaluation of posterior spinal fusion implants from various vendors and different spine levels. Eur Radiol 22:235723642012

    • Search Google Scholar
    • Export Citation
  • 7

    Hemmingsson AJung BYtterbergh C: Dual energy computed tomography: simulated monoenergetic and materialselective imaging. J Comput Assist Tomogr 10:4904991986

    • Search Google Scholar
    • Export Citation
  • 8

    Hibbs RA: An operation for progressive spinal deformities: a preliminary report of three cases from the service of the orthopaedic hospital. 1911. Clin Orthop Relat Res 460:17202007

    • Search Google Scholar
    • Export Citation
  • 9

    Joseph PMSpital RD: The effects of scatter in x-ray computed tomography. Med Phys 9:4644721982

  • 10

    Knott PTMardjetko SMKim RHCotter TMDunn MMPatel ST: A comparison of magnetic and radiographic imaging artifact after using three types of metal rods: stainless steel, titanium, and vitallium. Spine J 10:7897942010

    • Search Google Scholar
    • Export Citation
  • 11

    Lee MJKim SLee SASong HTHuh YMKim DH: Overcoming artifacts from metallic orthopedic implants at high-field-strength MR imaging and multi-detector CT. Radiographics 27:7918032007

    • Search Google Scholar
    • Export Citation
  • 12

    Lee YHPark KKSong HTKim SSuh JS: Metal artefact reduction in gemstone spectral imaging dual-energy CT with and without metal artefact reduction software. Eur Radiol 22:133113402012

    • Search Google Scholar
    • Export Citation
  • 13

    Liu PTPavlicek WPPeter MBSpangehl MJRoberts CCPaden RG: Metal artifact reduction image reconstruction algorithm for CT of implanted metal orthopedic devices: a work in progress. Skeletal Radiol 38:7978022009

    • Search Google Scholar
    • Export Citation
  • 14

    Mori IMachida YOsanai MIinuma K: Photon starvation artifacts of X-ray CT: their true cause and a solution. Radiol Phys Technol 6:1301412013

    • Search Google Scholar
    • Export Citation
  • 15

    Schenzle JCSommer WHNeumaier KMichalski GLechel UNikolaou K: Dual energy CT of the chest: how about the dose?. Invest Radiol 45:3473532010

    • Search Google Scholar
    • Export Citation
  • 16

    Srinivasan AHoeffner EIbrahim MShah GVLaMarca FMukherji SK: Utility of dual-energy CT virtual keV monochromatic series for the assessment of spinal transpedicular hardware-bone interface. AJR Am J Roentgenol 201:8788832013

    • Search Google Scholar
    • Export Citation
  • 17

    Valentin J: International Commission on Radiation Protection: Managing patient dose in multi-detector computed tomography (MDCT). ICRP publication 102. Ann ICRP 37:179iii2007

    • Search Google Scholar
    • Export Citation
  • 18

    Yu LLeng SMcCollough CH: Dual-energy CT-based monochromatic imaging. AJR Am J Roentgenol 199:5 SupplS9S152012

If the inline PDF is not rendering correctly, you can download the PDF file here.

Article Information

Contributor Notes

Correspondence Max Wintermark, Department of Radiology and Medical Imaging, Division of Neuroradiology, University of Virginia, Box 800170, Charlottesville, VA 22908-0170. email: max.wintermark@gmail.com.INCLUDE WHEN CITING Published online November 7, 2014; DOI: 10.3171/2014.10.SPINE14463.DISCLOSURE Dr. Smith is a consultant for Biomet, NuVasive, Medtronic, and DePuy. He reports receiving support of non–study-related clinical or research from DePuy/ISSG. Dr. Shaffrey is a consultant for Biomet, Globus, Medtronic, and NuVasive. He owns stock in NuVasive. He also is a patent holder with and receives royalties from Biomet, Medtronic, and NuVasive.
Headings
Figures
  • View in gallery

    Photograph of the anthropomorphic phantom in the gantry of the CT scanner. The metallic rod was inserted into the spinal access channel of the phantom (arrow), which was then filled with water and taped. Figure is available in color online only.

  • View in gallery

    Representative montage of axial and sagittal dual-energy CT images of the cervical spine with titanium (Ti) or cobalt-chromium (Co-Cr) alloy fixation rod inserted in the spinal canal of the anthropomorphic phantom with settings indicated on the right (medium or large FOV, 20 and 40-mm collimation), reconstructed at polychromatic (Poly) and calculated monoenergetic values of 70, 90, 110, and 130 keV. Lookup table window center and width were 40 and 350 HU, respectively.

  • View in gallery

    A: Representative graph of voxel intensities along the circular line profile around the spine (see circle in inset). B: Graph depicting the variability of voxel intensities (mean ± SD), an index of artifact conspicuity, as a function of reconstructed x-ray energy, FOV, and collimation when either titanium or cobalt-chromium alloy fixation rods were placed in the spinal canal.

References
  • 1

    Albee FH: Transplantation of a portion of the tibia into the spine for Pott's disease: a preliminary report 1911. Clin Orthop Relat Res 460:14162007

    • Search Google Scholar
    • Export Citation
  • 2

    Archer BRWhitmore RCNorth LBBushong SC: Bone marrow dose in chest radiography: the posteroanterior vs. anteroposterior projection. Radiology 133:2112161979

    • Search Google Scholar
    • Export Citation
  • 3

    Bamberg FDierks ANikolaou KReiser MFBecker CRJohnson TR: Metal artifact reduction by dual energy computed tomography using monoenergetic extrapolation. Eur Radiol 21:142414292011

    • Search Google Scholar
    • Export Citation
  • 4

    Bauer RWKramer SRenker MSchell BLarson MCBeeres M: Dose and image quality at CT pulmonary angiography-comparison of first and second generation dualenergy CT and 64-slice CT. Eur Radiol 21:213921472011

    • Search Google Scholar
    • Export Citation
  • 5

    Grams AESender JMoritz RObert MStein MOertel M: Dual energy CT myelography after lumbar osteosynthesis. Rofo 186:6706742014

    • Search Google Scholar
    • Export Citation
  • 6

    Guggenberger RWinklhofer SOsterhoff GWanner GAFortunati MAndreisek G: Metallic artefact reduction with monoenergetic dual-energy CT: systematic ex vivo evaluation of posterior spinal fusion implants from various vendors and different spine levels. Eur Radiol 22:235723642012

    • Search Google Scholar
    • Export Citation
  • 7

    Hemmingsson AJung BYtterbergh C: Dual energy computed tomography: simulated monoenergetic and materialselective imaging. J Comput Assist Tomogr 10:4904991986

    • Search Google Scholar
    • Export Citation
  • 8

    Hibbs RA: An operation for progressive spinal deformities: a preliminary report of three cases from the service of the orthopaedic hospital. 1911. Clin Orthop Relat Res 460:17202007

    • Search Google Scholar
    • Export Citation
  • 9

    Joseph PMSpital RD: The effects of scatter in x-ray computed tomography. Med Phys 9:4644721982

  • 10

    Knott PTMardjetko SMKim RHCotter TMDunn MMPatel ST: A comparison of magnetic and radiographic imaging artifact after using three types of metal rods: stainless steel, titanium, and vitallium. Spine J 10:7897942010

    • Search Google Scholar
    • Export Citation
  • 11

    Lee MJKim SLee SASong HTHuh YMKim DH: Overcoming artifacts from metallic orthopedic implants at high-field-strength MR imaging and multi-detector CT. Radiographics 27:7918032007

    • Search Google Scholar
    • Export Citation
  • 12

    Lee YHPark KKSong HTKim SSuh JS: Metal artefact reduction in gemstone spectral imaging dual-energy CT with and without metal artefact reduction software. Eur Radiol 22:133113402012

    • Search Google Scholar
    • Export Citation
  • 13

    Liu PTPavlicek WPPeter MBSpangehl MJRoberts CCPaden RG: Metal artifact reduction image reconstruction algorithm for CT of implanted metal orthopedic devices: a work in progress. Skeletal Radiol 38:7978022009

    • Search Google Scholar
    • Export Citation
  • 14

    Mori IMachida YOsanai MIinuma K: Photon starvation artifacts of X-ray CT: their true cause and a solution. Radiol Phys Technol 6:1301412013

    • Search Google Scholar
    • Export Citation
  • 15

    Schenzle JCSommer WHNeumaier KMichalski GLechel UNikolaou K: Dual energy CT of the chest: how about the dose?. Invest Radiol 45:3473532010

    • Search Google Scholar
    • Export Citation
  • 16

    Srinivasan AHoeffner EIbrahim MShah GVLaMarca FMukherji SK: Utility of dual-energy CT virtual keV monochromatic series for the assessment of spinal transpedicular hardware-bone interface. AJR Am J Roentgenol 201:8788832013

    • Search Google Scholar
    • Export Citation
  • 17

    Valentin J: International Commission on Radiation Protection: Managing patient dose in multi-detector computed tomography (MDCT). ICRP publication 102. Ann ICRP 37:179iii2007

    • Search Google Scholar
    • Export Citation
  • 18

    Yu LLeng SMcCollough CH: Dual-energy CT-based monochromatic imaging. AJR Am J Roentgenol 199:5 SupplS9S152012

TrendMD
Cited By
Metrics

Metrics

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 564 401 16
PDF Downloads 226 49 5
EPUB Downloads 0 0 0
PubMed
Google Scholar