Introduction
-
To determine the accuracy of the EOS imaging system compared to CT for measuring native hip/pelvic parameters.
-
To determine the accuracy of the EOS imaging system compared to CT for measuring postoperative/prosthetic hip parameters.
Methods
Search strategy
Study selection
Inclusion criteria
Exclusion criteria
Data extraction
Quality assessment
Strategy for data synthesis
Results
Search strategy
Study characteristics
Study | Author (year) | Study design and recruitment date | Sample size (n) | Mean age (years) | Reference test (CT scan) | EOS imaging protocol | Main findings |
---|---|---|---|---|---|---|---|
1 | Buck, F. M., et al. (2012) | Not reported | 35 patients (12 male) | Women: 64 range 46–89 Men: 67 range 57–78 | 40- MDCT scanner (Philips) axial images (140 kV; 300mAs; matrix 512 × 512; pitch .426; .45 mm reconstruction increment; 1 mm reconstruction thickness) | 40 kV and 250 mAs | High correlation for all measurements with P < 0.001 High correlation with ICC of 0.952 (95% CI, 0.905–0.976) and 0.938 (95% CI, 0.878–0.969) with CT, for femoral torsion and tibial torsion respectively High correlation with ICC of 0.943 (95% CI, 0.886–0.971) and 0.959 (95% CI, 0.918–0.979) with EOS, for femoral torsion and tibial torsion respectively High correlation for femoral anteversion (AV) angle with IC = 0.952 |
2 | Folinais, D., et al. (2013) | Retrospective study (between November 2009 and March 2011 | 43 lower limbs/ 30 patients (15 male) | 53.2 ± 20.4 | Helicat CT scanner/ 40-slice | Not reported | Mean difference of 6.3° and 6.8° between EOS and CT, respectively; for femoral torsion, with P = 0.6 Mean difference of 3.9° for tibial torsion High correlation with ICC with EOS 0.93 and 0.86 for femoral and tibial torsion, respectively; ICC with CT was 0.90 and 0.92 for femoral and tibial torsion, respectively For anteroposterior in EOS, radiation dose was 0.18 ± 0.05 mGy, and 0.45 ± 0.08 mGy for lateral position For CT, radiation dose was 8.4 to 15.6 mGy High correlation with ICC approximately 0.9 for both EOS and CT High correlation between EOS and CT with r = 0.93 for femoral anteversion (AV) angle |
3 | Tokunaga, K., et al. (2018) | Not reported | 90 patients (15 male) | 60 | 80-slice CT machine (Aquilion PRIME, Toshiba) | Not reported | High correlation between EOS and CT for implant angle measurements of THA, with correlation coefficients (rho) as follows: stem antetorsion = 0.8861, cup anatomical anteversion = 0.763 and cup radiographic inclination = 0.679, with P < 0.0001 High correlation between EOS and CT in measurement values, but a difference of 5° in cup anteversion and stem antetorsion due to outliers |
4 | Fritz, Benjamin et al. (2019) | Retrospective study | 50 patients (29 male) | 69.7 range 53–87 | 64-slice CT scanner (Philips) 120 kV/250 mAs/collimation (64 × 0.625 mm)/ rotation time (0.5 s.)/1 mm slice thickness for axial images) | AP image (83–95 kV and 200-280 mA)/ For lateral image (102-120 kV and 200-320 mA) | High correlation between EOS and CT with ICC ≥ 0.8 for 2D and 3D acetabular coverage For global 3-D acetabular coverage measurements, difference between CT and BPR of only 0.9% with standard deviations of 3.6% and 3.0%, respectively |
5 | Esposito, Christina I., et al. (2020) | Prospective study | 20 patients | Not reported | Not reported | Not reported | Mean difference values between EOS and CT: 4° ± 4° for femoral anteversion, 3° ± 2° for acetabular anteversion and 2° ± 2° for acetabular inclination With CT, inter-rater correlation greater than 0.78 for all hip angle measurements; slight difference with EOS may be due to outliers High correlation with EOS as follows: For acetabular inclination (Cronbach’s α = 0.83) For acetabular anteversion (Cronbach’s α = 0.89) For femoral components (Cronbach’s α = 0.89) |
6 | Mayr, Hermann O., et al. (2021) | Observational study | 34 femora measured/ 19 patients (4 male) | 45.5 ± 19.8 | Not reported | Not reported | Both hip measurements (15 patients) One side measurement (four patients) Eleven hips had no torsional malalignment; fourteen had reduced anteversion (less than ten degrees) or retroversion (less than zero degrees) Nine hips had increased anteversion (more than 20 degrees) High correlation between for the assessment of femoral anteversion angle (AV), r = 0.855 in patients with regular AV (n = 34) Moderate correlation between (r = 0.495) in patients with reduced AV (less than 10 degrees) involving retroversion (less than zero degrees), (n = 14) Low correlation r = 0.292, in patients with increased AV (more than 20 degrees), (n = 9) Significant correlation in all measurements (n = 34; P < 0.001), in physiological AV (n = 11; P = 0.001), in decreased AV and retroversion (n = 14; P = 0.072) and in increased AV (n = 9; P = 0.446) High agreement on AV measurements between all 3 examiners with ICC = 0.911 using EOS and ICC = 0.934 using CT Good inter-observer reliability with Cronbach’s α values of 0.955, 0.934 for EOS and CT, respectively No correlation observed in patients with torsional malalignment |
Quality assessment
Study | Risk of bias (1 = low, 2 = medium, 3 = high) | Applicability concerns | ||||||
---|---|---|---|---|---|---|---|---|
Patient selection | Index test | Reference standard | Flow and timing | Patient selection | Index test | Reference standard | Flow and timing | |
Buck, F. M., et al. (2012) | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 |
Folinais, D., et al. (2013) | 1 | 1 | 1 | 2 | 1 | 1 | 1 | 1 |
Tokunaga, K., et al. (2018) | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 |
Fritz, Benjamin et al. (2019) | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 |
Esposito, Christina I., et al. (2020) | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 |
Mayr, Hermann O., et al. (2021) | 1 | 1 | 1 | 2 | 1 | 1 | 1 | 1 |