Diagnostic Performance of Dual-Energy CT for Detecting Painful Hip Prosthesis Loosening
Abstract
Background
Revisions of hip prostheses are increasing, and conventional radiography (CR) is a primary tool for managing complications. However, dual-energy CT (DECT) with virtual monoenergetic imaging is capable of reducing periprosthetic metal artifacts compared with standard CT.
Purpose
To compare the diagnostic performance of DECT and CR in detecting hip prosthesis loosening, using surgery as a reference for diagnosis.
Materials and Methods
This retrospective single-center study conducted between January 2018 and October 2020 included consecutive patients with unilateral painful hip prostheses. Two independent readers (with 15 years and 4 years of experience) who were blinded to clinical findings evaluated CR and DECT images. At imaging, diagnosis of loosening prosthesis was made for periprosthetic radiolucency greater than or equal to 2 mm wide or the presence of two or more secondary findings, including periprosthetic osteolysis, angulation of the implant, fracture, or abnormal periosteal reaction. For each reader and for each imaging parameter, sensitivity and specificity were calculated. The diagnostic performance of each imaging tool was compared by using the McNemar test. Interobserver agreements were calculated with Cohen κ statistics. Statistical software was used.
Results
Overall, 178 patients (mean age ± standard deviation, 74 years ± 20; 96 men) were included (121 undergoing surgery, 57 follow-up). Overall, 87 of 178 patients (49%) were diagnosed with a loosened prosthesis. DECT had higher sensitivity and specificity than CR for both reader 1 (94% [82 of 87 examinations; 95% CI: 87, 98] and 93% [85 of 91 examinations; 95% CI: 86, 97] vs 84% [73 of 87 examinations; 95% CI: 74, 91] and 91% [83 of 91 examinations; 95% CI: 83, 96], respectively; P < .001) and reader 2 (92% [80 of 87 examinations; 95% CI: 84, 97] and 95% [86 of 91 examinations; 95% CI: 88, 98] vs 80% [70 of 87 examinations; 95% CI: 71, 88] and 91% [83 of 91 examinations; 95% CI: 83, 96], respectively; P = .001), with better interobserver agreement (κ, 0.88 [95% CI: 0.81, 0.95] vs 0.78 [95% CI: 0.69, 0.88]).
Conclusion
Dual-energy CT showed better diagnostic performance than conventional radiography in diagnosing hip prosthesis loosening.
See also the editorial by Lutz in this issue.
© RSNA, 2021
Summary
Dual-energy CT with virtual monoenergetic high–kilovoltage peak reconstructions showed better diagnostic performance than conventional radiography in the assessment of hip prosthesis loosening.
Key Results
■ Among 178 patients with unilateral painful hip prostheses examined by using dual-energy CT (DECT) and conventional radiography (CR), 87 (49%) had a loosened prosthesis.
■ A periprosthetic gap of 2 mm or greater (P < .001) was associated with a loosened prosthesis.
■ DECT had higher sensitivity and specificity than CR, with better interobserver agreement for DECT (κ, 0.88 vs 0.78): reader 1 had a sensitivity of 94% and specificity of 93% with DECT versus 84% and 91% with CR (P < .001), and reader 2 had a sensitivity of 92% and specificity of 94% with DECT versus 80% and 91% with CR (P = .001).
Introduction
Hip arthroplasty with total prosthetic replacement is the reference standard for the treatment of symptomatic osteoarthritic degeneration of the hip joint (1). Over 1 million hip replacement surgeries are performed per year worldwide (2). Due to the aging population globally, an increase in the number of hip arthroplasties is anticipated (3,4). Hip replacement is an elective orthopedic procedure with a mean revision-free period of 10–15 years or even up to 25 years (5,6). Due to the increasing number of patients requiring treatment, revision surgery may become a frequent event in the management of patients with a prosthetic hip. The reasons for revision surgery vary, but aseptic loosening is the most frequent (4). Other reasons for revision surgery are prosthetic dislocation, instability, periprosthetic fractures, and infection (5).
Conventional radiography (CR) remains the standard imaging tool for the evaluation of hip implants (7–11). However, CR may be limited by the use of two-dimensional images. In particular, in a published study (12), CR demonstrated a sensitivity and specificity of 26% and 100%, respectively, for acetabular component loosening and 20% and 100% for femoral component loosening. Multidetector CT has a primary role in the management of prosthesis-related complications when radiographs are insufficient for a diagnosis. CT is more accurate for evaluating implant components, determining prosthesis geometry and position, and assessing the periprosthetic bone (13).
Despite its extensive use, projection CT is limited by the presence of metal-induced artifacts. These artifacts, presenting as alternating markedly hypo- or hyperattenuating bands on the image, are caused by photon starvation, beam hardening, and scattering (13).
Dual-energy CT (DECT) allows virtual monoenergetic imaging (MEI) reconstructions at a high kilovoltage peak. Several studies have demonstrated that these reconstructions, both alone and in combination with iterative metal artifact reconstruction, or IMAR, can improve image quality by reducing metal artifacts in the periprosthetic bone and soft tissues (12–17). However, there are few articles describing the role of DECT in the evaluation of loosening hip prostheses. In one recent article, Guziński et al (15) analyzed the efficacy of DECT monochromatic reconstructions in the diagnosis of loosening hip protheses among 25 consecutive patients. In that study, high-energy monochromatic images proved superior to low-energy DECT with iterative metal artifact reconstruction images in assessing loosening implants. The purpose of our study was to compare the diagnostic performance of DECT and CR in the identification of loosening hip prostheses.
Materials and Methods
Patients
This retrospective single-center study (project number 2765 CESC) was approved by the institutional review board, and informed consent was obtained from all patients enrolled. Between January 2018 and October 2020, consecutive patients were considered for inclusion. Eligibility criteria were the presence of unilateral painful hip prosthesis at orthopedic visit, diagnosis of loosening by means of surgical revision within 14 days of imaging studies, or diagnosis of nonloosening prosthesis as confirmed by means of 1 year clinical-radiologic follow-up. Exclusion criteria included concomitant oncologic disease, lack of DECT scan, lack of surgical confirmation of prosthesis loosening, and intraoperative signs of an infected implant.
CR Protocol
Preoperative radiographic evaluation consisted of a standard anteroposterior view of the pelvis and frog-leg and/or cross-table lateral views of the affected hip (10).
DECT Protocol
Noncontrast DECT examinations were performed with a 384-section dual-source CT scanner (Somatom Definition Force, Siemens Healthineers). The scanning parameters were tube voltages of 80/150 kV, with tin filter, and predefined tube current–time product of 1.6:1 (tube A, 220 mAs; tube B, 138 quality reference mAs) with detector collimation at 32 × 0.6 mm and pitch at 0.6. Automated attenuation-based tube current modulation (CARE Dose 4D, Siemens Healthcare) was used.
The mean effective postscan CT dose index volume was 6.64 mGy (range ± standard deviation, 5.4–8.7 mGy ± 0.40); the mean dose-length product was 268.23 mGy · cm (range ± standard deviation, 234.3–294.7 mGy · cm ± 21.20).
DECT Postprocessing
Soft-tissue kernel (Qr32) 80-kVp and 150-kVp set images (thickness, 0.75 mm; increment, 0.6 mm) were transferred to an off-line workstation (Syngo.Via VB20, Siemens Healthineers).
By using MEI plus application, depending on a visual qualitative assessment, the readers were free to choose the kilovoltage peak value yielding the best imaging quality. For each patient, the isotropic image data set was analyzed with bone window on the preferred imaging planes.
Image Analysis
CR and MEI plus application images were independently analyzed in a random order by two radiologists (G.F. [reader 1] and A.F. [reader 2], with 15 years and 4 years of experience in musculoskeletal radiology and 6 years and 1 year of experience with DECT imaging, respectively) blinded to clinical findings, with a 2-month interval between reading sessions.
At imaging, the diagnosis of a loosening prosthesis was based on the presence of a direct sign of loosening (radiolucency ≥2 mm wide), or in cases of two or more indirect findings, a binary approach was used, distinguishing presence from absence.
Periprosthetic radiolucency was determined as a low-density elongated strip between the cement mantle and bone, or the radiolucency around the components, measured in millimeters on any plane and in any location at the level of both the femoral stem and the acetabular component of the prosthesis. A 2-mm width cutoff was used (radiolucency ≥2 mm was considered a diagnosis of loosening, independently from indirect findings; radiolucency <2 mm was considered an absence of loosening) (7–10). Indirect findings were evaluated on the basis of visual assessment and considered for diagnosis in the absence of detectable radiolucency (10). Indirect findings included periprosthetic osteolysis, defined as focal areas of bone reabsorption or asymmetric areas of reduced bone conspicuity near the implant; abnormal angulation or dislocation of the femoral stem, defined as the loss of neutral alignment with the longitudinal axis of the femoral shaft; abnormally increased acetabular inclination angle; and any fracture or abnormal periosteal reaction adjacent to the implant (10).
Surgery and Follow-up
In the subgroup of patients undergoing surgery, the diagnosis of loosening prosthesis was confirmed intraoperatively by the surgeon with use of dedicated procedures, distinguishing acetabular from femoral stem loosening. However, a direct correlation between images and surgery was not performed. In the subgroup of patients not undergoing surgery, a 1-year clinical-radiologic follow-up was used to confirm the absence of loosening phenomena. Body mass index values were assessed for all patients and correlated with the final diagnosis.
Statistical Analysis
Continuous variables (age, body mass index, and periprosthetic radiolucency at imaging [measured in millimeters]) were summarized with means, standard deviations, and ranges, while count variables (periprosthetic osteolysis, angulation, and any fracture adjacent to the implant) were summarized with absolute and percentage frequencies. Statistical models and estimations were adjusted for covariates as necessary. For the diagnostic performance, tests results were displayed in contingency tables. For each reader and each modality, sensitivity and specificity were calculated. Estimations were reported with exact binomial 95% CIs. The intraoperative findings or the findings at 1-year follow-up were used as the reference standard for the diagnosis.
Interrater reliability was estimated with use of the Cohen κ index and its 95% asymptotic CI. Four multivariable logistic regression models (for each reader and each imaging tool) were used to predict the probability of loosening prosthesis and adjusted for the following predictors: periprosthetic radiolucency (presence vs absence), secondary findings (presence vs absence), body mass index, age, and sex. A comparison between DECT and CR performance values was carried out by using the McNemar test. P < .05 was considered indicative of statistically significant difference. StataCorp statistical software (version 15) was used.
Results
Patient Characteristics
Overall, 191 patients were considered for inclusion. Thirteen patients were excluded because of concomitant oncologic disease (n = 2), lack of DECT scan (n = 3), lack of surgical confirmation of prosthesis loosening (n = 5), or intraoperative signs of septic loosening (n = 3). After exclusions, there was a total of 178 patients (mean age, 74 years; range, 22–90 years; 96 men [54%]). Among the patients enrolled, 121 underwent revision surgery, whereas 57 did not. The patients’ clinical data are summarized in Table 1. Figure 1 is a flow diagram showing patients included in the study. The tabulation of imaging parameters and the diagnostic performance values are described in Tables 2 and 3, respectively.
 |
Figure 1: Flow diagram shows initial number of participants and those excluded for any reason. CR = conventional radiography, DECT = dual-energy CT.
 |
 |
At surgery, a diagnosis of a loosened prosthesis was established in 87 of the 178 patients (49%; 36 with acetabular, 42 with femoral stem, and nine with acetabular and femoral stem loosening). Among the 34 surgical patients without intraoperative signs of loosening, the diagnosis was implant rupture or mechanical failure (n = 21), metallosis (n = 6), abnormal heterotopic bone formation (n = 3), periprosthetic lipoma (n = 2), or periprosthetic hematoma (n = 2).
Reader Performances
Reader 1 measured a periprosthetic radiolucency of 2 mm or greater in 81 of 178 DECT examinations (46%) (mean value ± standard deviation, 2.6 mm ± 3.7) and in 78 of 178 CR studies (44%) (mean value, 2.4 mm ± 3.4); reader 2 measured a periprosthetic radiolucency of 2 mm or greater in 77 of 178 DECT examinations (43%) (mean value, 2.4 mm ± 3.7) and in 74 of 178 CR studies (42%) (mean value, 2.4 mm ± 3.6).
Reader 1 identified two or more secondary signs of loosening in 54 of 178 examinations (30%) at DECT, with two false-positive results, and in 49 of 178 examinations (28%) at CR, with two false-positive results. Reader 2 identified two or more indirect signs of loosening in 51 of 178 examinations (29%) at DECT, with one false-positive result, and in 48 of 178 examinations (27%) at CR, with two false-positive results.
Overall, reader 1 correctly diagnosed 82 of 87 loosened prostheses at DECT (34 patients with acetabular, 40 with femoral stem, and eight with acetabular and femoral stem loosening) and 73 of 87 loosened prostheses at CR (30 patients with acetabular, 35 with femoral stem, and eight with acetabular and femoral stem loosening), with sensitivity and specificity of 94% (95% CI: 87, 98) and 93% (95% CI: 86, 97), respectively, for DECT and 84% (95% CI: 74, 91) and 91% (95% CI: 83, 96) for CR.
Overall, reader 2 correctly diagnosed 80 of 87 loosened prostheses at DECT (34 patients with acetabular, 38 with femoral stem, and eight with acetabular and femoral stem loosening) and 70 of 87 loosened prostheses at CR (28 patients with acetabular, 34 with femoral stem, and eight with acetabular and femoral stem loosening), with sensitivity and specificity of 92% (95% CI: 84, 97) and 94% (95% CI: 88, 98), respectively, for DECT and 80% (95% CI: 71, 88) and 91% (95% CI: 83, 96) for CR.
Loosening was missed in 12 cases at DECT—five by reader 1 and seven by reader 2, with three overlapping cases, two missed by reader 1 alone, and four missed by reader 2 alone. In two of these patients, a small periacetabular radiolucency of 2 mm was found at the postsurgical revision of DECT images, whereas no findings were detected at CR. Reader 1 missed loosening in 14 cases at CR and reader 2 missed 17 cases, with 10 overlapping cases, four missed by reader 1 alone and seven missed by reader 2 alone. Figures 2–5 show example images in patients in which at least one reader missed loosening at one or more modalities.
Figure 2: Images in a 72-year-old man with right acetabular loosening hip implant at surgical revision. (A) Coronal and (B) sagittal noncontrast reconstructed 1-mm dual-energy CT monoenergetic imaging plus application scans show periacetabular radiolucency of 2 mm or greater (arrows). On the corresponding (C) anteroposterior and (D) frog-leg conventional radiographs, the radiolucency (arrow in D) was confirmed along the lateral aspect of the acetabular component by reader 1 only. In addition, a fractured acetabular screw fixation can be recognized (arrow in C).
Figure 3: Images in a 65-year-old woman with a left acetabular loosening hip implant at surgical revision. On coronal (A) and axial (B) noncontrast reconstructed 1-mm dual-energy CT (DECT) monoenergetic imaging plus application scans, only a small focal radiolucency (arrow in A and B) can be recognized close to the acetabular component. On the corresponding (C) anteroposterior and (D) frog-leg radiographs, there is no evidence of any periacetabular radiolucency (arrow). The case was misdiagnosed as a false-negative result by both readers at DECT and conventional radiography.
Figure 4: Images in a 65-year-old man with left femoral stem loosening hip implant at surgical revision. Coronal (A) and axial (B) reconstructed 1-mm noncontrast dual-energy CT monoenergetic imaging plus application scans show the presence of a large radiolucency between the lateral aspect of femoral stem and the greater trochanter (arrow). (C, D) On the corresponding anteroposterior radiographs, the periprosthetic radiolucency (arrow) was missed by both readers.
Figure 5: Images in a 71-year-old man with left loosening hip implant at surgical revision show both primary and secondary signs of prosthesis loosening. (A) Coronal noncontrast reconstructed 1-mm dual-energy CT (DECT) monoenergetic imaging plus application scan shows a subtle periacetabular area of bone reabsorption (arrow) and periprosthetic radiolucency with periosteal reaction (arrowheads). (B) On the corresponding anteroposterior radiograph, neither reader pointed out any substantial periprosthetic radiolucency. (C) On the parasagittal 1-mm DECT image, the radiolucency and periosteal reaction are confirmed, and a small fracture (arrow) is seen on the lateral margin of the femur. (D) On the corresponding frog-leg radiograph, a periprosthetic radiolucency (arrowheads) along the femoral stem can be seen on both medial and lateral aspect.
At DECT, the interobserver agreement for the diagnosis of loosened prostheses was near-perfect (κ, 0.88; 95% CI: 0.81, 0.95), whereas at CR, agreement was good (κ, 0.78; 95% CI: 0.69, 0.88).
When the McNemar test was applied, a statistically significant difference was determined between the overall diagnosis of loosening prosthesis at DECT and CR for reader 1 (P < .001) and reader 2 (P = .002) and for the presence of periprosthetic radiolucency for reader 1 (P = .04) and the identification of secondary findings for reader 1 (P = .03).
The diagnostic performance of DECT and CR was not influenced by the site of loosening (acetabular vs femoral) for reader 1 (P = .18 and .78, respectively) or reader 2 (P = .70 and .57, respectively).
For the MEI plus application images, a mean value of 155 kVp (range, 140–165 kVp) was chosen by the readers.
Multivariable Logistic Regression Model of Primary and Secondary Imaging Findings
The results of the multivariate logistic model are shown in Table 4. The presence of loosening at surgery was not correlated with the body mass index, age, or sex of patients enrolled. The presence of loosening at surgery was associated with the presence of periprosthetic radiolucency at DECT, as detected by reader 1 (P = .004); with the presence of periprosthetic radiolucency at CR, as detected by reader 1 (P = .03) and reader 2 (P = .03); and with the presence of secondary findings at CR, as detected by reader 1 (P = .004) and reader 2 (P = .001).
 |
Discussion
Revision surgery procedures of hip prostheses are increasing worldwide, representing a common clinical problem (3,4). Conventional radiography (CR) is the primary tool for diagnosing loosening, but it may be limited by the use of two-dimensional images, with subsequent low sensitivity values (12). Dual-energy CT (DECT) is capable of reducing periprosthetic metal artifacts compared with standard CT, allowing fine evaluation of the bone-prosthesis interface. In our study, DECT yielded higher sensitivity and specificity as compared with CR in diagnosing loosening phenomena for both reader 1 (94% and 93%, respectively, for DECT vs 84% and 91%, respectively, for CR; P < .001) and reader 2 (92% and 94% vs 80% and 91%; P = .001). In particular, DECT helped reader 1 to better identify periprosthetic radiolucency and reader 2 to identify the secondary findings. Moreover, the interobserver agreement for DECT was superior to that for CR (κ, 0.88 vs 0.78, respectively).
In a meta-analysis including 32 published articles (12), the diagnostic performance of CR in the detection of aseptic loosening of the femoral component was 82% sensitivity and 81% specificity. In another study (11), MRI had a sensitivity and specificity of 83% and 98%, respectively, for acetabular component loosening and 75% and 100%, respectively, for femoral component loosening. Radiography had a sensitivity and specificity of 26% and 100%, respectively, for acetabular component loosening and 20% and 100%, respectively, for femoral component loosening.
In our study, CR sensitivity values (84% and 80% for readers 1 and 2, respectively) were in line with the mean value (82%) reported by Temmerman et al (12). Conversely, the specificity values (93% and 91% for readers 1 and 2, respectively) were superior to that described in the aforementioned study (81%) but inferior to that reported by Burge et al (100%) (11). In this regard, when a radiolucency is clearly depicted at CR, the work-up could be considered concluded. Conversely, DECT should be used in case of inconclusive CR, especially for patients with high pain levels.
Although several approaches were suggested to reduce the artifacts caused by prostheses, including the optimization of single-energy CT scanning (14), iterative reconstructions, and specific software for iterative metal artifact reduction (18,19), these techniques can generate new artifacts, such as blurring and shape distortion (20–26). Moreover, in the study by Bamberg et al (23), image quality with DECT was superior to that with single-energy CT in 29 of 31 cases. Furthermore, Lewis et al (24) demonstrated a reduction of streak artifacts produced by metallic prostheses without deterioration in contrast when using high–kiloelectron volt monoenergetic reconstruction (21). For these reasons, in our study, only MEI was used for diagnosis (20,23–26).
In line with the results of Guziński et al (15), our study has confirmed a role for MEI scans in predicting the loosening of metallic hip implants. The majority of loosened prostheses showed a sizeable periprosthetic radiolucency at DECT and CR.
In cases of loosening without any measurable radiolucency, additional secondary findings could play an important role. However, the secondary findings were characterized by high specificity but relatively low sensitivity values. As a consequence, in the absence of any discernible radiolucency, the diagnosis of a loosening hip implant in our study was based on the presence of at least two indirect signs, which applied to five and eight patients at DECT for readers 1 and 2, respectively, and two patients per reader for CR. However, we emphasize that secondary findings could assist diagnosis by increasing sensitivity of imaging at both DECT and CR.
According to the multivariable logistic regression model, the diagnosis of loosening was not influenced by the body mass index, age, or sex of the patients enrolled. Conversely, the diagnosis of loosening was positively correlated with the presence of a periprosthetic radiolucency of 2 mm or greater and periprosthetic osteolysis at DECT and with the presence of secondary findings at CR.
Our study had some limitations, including a lack of DECT image quality assessment. However, the role of DECT in reducing metal-induced artifacts has been described in previously published research. We focused on the widest point of the periprosthetic radiolucency, without considering its length or creating a zonal scoring system. A direct zonal correlation between imaging and surgery was also not carried out. During surgical revision, the orthopedic surgeons could only confirm or rule out the loosening of the acetabular or femoral components. Furthermore, data regarding osteopenia were not available for analysis.
In conclusion, dual-energy CT (DECT) with high–kilovoltage peak monoenergetic reconstructions was more accurate than conventional radiography in the diagnosis of hip prosthesis loosening. Larger studies in the future could help corroborate the role of DECT in the work-up of loosening prosthesis, especially in cases of inconclusive radiography and for surgical planning of difficult cases.
Acknowledgment
We thank Elinor Julie Rae Anderson, PhD, for English editing.
Author Contributions
Author contributions: Guarantors of integrity of entire study, G.F., A.F., M.G., C.Z., G.C.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; agrees to ensure any questions related to the work are appropriately resolved, all authors; literature research, G.F., A.F., A.C., S.N., C.Z.; clinical studies, A.F., A.C., S.N., C.Z., G.C.; statistical analysis, G.F., A.F., A.C., S.N., M.G., C.Z.; and manuscript editing, G.F., A.F., A.C., S.N., C.Z., G.C.
References
- 1. . Periprosthetic osteolysis: mechanisms, prevention and treatment. J Clin Med 2019;8(12):2091.
- 2. . Hip replacement. Lancet 2018;392(10158):1662–1671.
- 3. . Effects of an aging population on hospital costs related to elective hip replacements. Public Health 2020;180(10):16.
- 4. . Global perspectives on arthroplasty of hip and knee joints. Orthop Clin North Am 2020;51(2):169–176.
- 5. . Twenty-five-year survivorship of two thousand consecutive primary Charnley total hip replacements: factors affecting survivorship of acetabular and femoral components. J Bone Joint Surg Am 2002;84(2):171–177.
- 6. . How long does a hip replacement last? A systematic review and meta-analysis of case series and national registry reports with more than 15 years of follow-up. Lancet 2019;393(10172):647–654.
- 7. . From the RSNA refresher courses. Total hip arthroplasty: radiographic evaluation. RadioGraphics 1996;16(3):645–660.
- 8. . Hip arthroplasty. Part 1: prosthesis terminology and classification. Clin Radiol 2009;64(10):954–960.
- 9. . Hip arthroplasty. Part 2: normal and abnormal radiographic findings. Clin Radiol 2009;64(10):961–971.
- 10. . Postoperative radiograph of the hip arthroplasty: what the radiologist should know. Insights Imaging 2015;6(6):591–600.
- 11. . Erratum to: What is the diagnostic accuracy of MRI for component loosening in THA? Clin Orthop Relat Res 2019;477(10):2387.
- 12. . Accuracy of diagnostic imaging techniques in the diagnosis of aseptic loosening of the femoral component of a hip prosthesis: a meta-analysis. J Bone Joint Surg Br 2005;87(6):781–785.
- 13. . CT of the hip prosthesis: appearance of components, fixation, and complications. RadioGraphics 2012;32(4):1089–1107.
- 14. . Reducing artifacts from total hip replacements in dual layer detector CT: combination of virtual monoenergetic images and orthopedic metal artifact reduction. Eur J Radiol 2019;111(14):20.
- 15. . Dual-energy computed tomography in loosening of revision hip prosthesis: a comparison between MARS and non-MARS images. J Comput Assist Tomogr 2019;43(3):379–385.
- 16. . Imaging of hip arthroplasties: normal findings and hardware complications. Semin Musculoskelet Radiol 2019;23(2):162–176.
- 17. . CT of hip prosthesis: new techniques and new paradigms. Diagn Interv Imaging 2016;97(7-8):725–733.
- 18. . CT radiation dose reduction in patients with total hip arthroplasties using model-based iterative reconstruction and orthopaedic metal artefact reduction. Skeletal Radiol 2019;48(11):1775–1785.
- 19. . Quantifying metal artefact reduction using virtual monochromatic dual-layer detector spectral CT imaging in unilateral and bilateral total hip prostheses. Eur J Radiol 2017;88(61):70.
- 20. . Metal artifacts reduction using monochromatic images from spectral CT: evaluation of pedicle screws in patients with scoliosis. Eur J Radiol 2013;82(8):e360–e366.
- 21. . The utility of digital linear tomosynthesis imaging of total hip joint arthroplasty with suspicion of loosening: a prospective study in 40 patients. BioMed Res Int 2013;2013594631.
- 22. . Artifacts in CT: recognition and avoidance. RadioGraphics 2004;24(6):1679–1691.
- 23. . Metal artifact reduction by dual energy computed tomography using monoenergetic extrapolation. Eur Radiol 2011;21(7):1424–1429.
- 24. . Reducing the effects of metal artefact using high keV monoenergetic reconstruction of dual energy CT (DECT) in hip replacements. Skeletal Radiol 2013;42(2):275–282.
- 25. . Application and advantages of monoenergetic reconstruction images for the reduction of metallic artifacts using dual-energy CT in knee and hip prostheses. Radiol Med (Torino) 2018;123(8):593–600.
- 26. . Monoenergetic imaging of dual-energy CT reduces artifacts from implanted metal orthopedic devices in patients with factures. Acad Radiol 2011;18(10):1252–1257.
Article History
Received: Aug 21 2020Revision requested: Sept 21 2020
Revision received: Apr 20 2021
Accepted: May 14 2021
Published online: July 06 2021
Published in print: Sept 2021
