MRI-based assessments of hip cartilage or volume in symptomatic (OA) or asymptomatic populations
| Authors | Demographics | MRI details | Image analysis and reliability | Hip cartilage measure(s) | Defining criteria for OA | Incidence and/or prevalence estimates |
|---|---|---|---|---|---|---|
| Cicuttini et al, 200028 | n=6: 3M, 3F. Ages 24–65 years. | 1.5T, fat-sat T1, side not specified, 3D volume calculation from sagittal images | FHCV determined estimation of isotropic voxel size by trilinear interpolation, manual contouring, and data resampling. FHCV estimated by summing pertinent voxels by one user. Intraobserver CoV for FHCV was 6.6%. ICC for FHCV was 0.94. | FHCV (1800–7800 mm3) | None | None |
| Zhai et al, 200529 | n=151: participants from TASOACS. 79 M, 72 F. Mean age 63 years | 1.5T, 3D fat-sat T1 GRE, right hip, sagittal images | Same technique as above Intraobserver and interobserver reliabilities CoV 2.5% and 4.4% | FHCV (M: 5900±1000 and F: 4700±800 mm3) FHCT (M: 1.6±0.2 mm F: 1.7±0.2 mm) | Altman | Radiographic OA in 46% of M and 56% of F. No prevalence estimates for MRI OA |
| Naish et al 200630 | n=6: all female. Ages 22–34 years | 1.5T, 3D gradient echo, right hip, sagittal images | WJCV segmentation to subvoxel accuracy using a semiautomated method. Intraobserver CoV for volume of knee cartilage 1.8% (data not shown). Further detail in Gougoutas et al 200431 | WJCT (2.3±0.13 mm) | None | None |
| Carballido-Gamio et al, 200832 | n=7: 5 asymptomatic, 2 with radiographic OA. Mean age 26.6±7.4 years for normal, 54 and 61 years for OA participants | 3T, T1ρ and T2 relaxometry, either hip, sagittal images | WJCV and WJCT segmentation by one user with a semiautomated technique. Correlation coefficient of volume and thickness estimates compared with saline displacement >0.95. Further detail in Carballido-Gamio et al 200533 | WJCV (6263 mm3) WJCT (3.12 mm) | K-L scale. K-L≤2 mild, K-L>2 advanced | None |
| Khan et al, 201334 | n=151: participants from TASOACS. 79M, 72F. Ages 50–81 years | 1.5T, fat-sat T1 GRE, right hip, 3D volume calculation from sagittal images | Identical methods to Zhai et al, 200529 | FHCV (5297 mm3) | Altman | None |
| Ahedi et al, 201435 | n=243: participants from TASOACS. M/F not specified. Mean age 64 years | Field strength not specified, STIR, right hip, plane not specified | BMLs identified as areas of increased signal intensity on STIR adjacent to the subchondral bone. One user assessed maximum area of the lesion by manual contouring. ICC was 0.98. | Semiquantitative assessment of hip cartilage defects | Altman | 77% had hip cartilage defects, not related to age, sex, or BMI |
| Teichtahl et al, 201410 | n=160: participants from MCCS. n=141 non-OA, 19 OA. 56F and 58M. Mean age 66.8 and 59.2 years, respectively | 3T, 3D GRE fat-sat T2, PD, spin echo, dominant hip, sagittal images | Identical methods to Zhai et al, 2005.29 ICC was 0.99. | FHCV (OA: 1763 mm3 Non-OA: 3343 mm3) | K-L scale | Prevalence of BML and cartilage defects in OA and non-OA participants |
| Teichtahl et al, 201511 | n=141: participants from MCCS. 62 M, 79 F. Non-OA. Mean age 66.8 years | 3T, 3D GRE fat-sat T2, PD, spin echo, dominant hip, sagittal images | Identical methods to Zhai et al, 2005.29 ICC was 0.99. | FHCV (M; 3891 mm3 F: 2867 mm3) | None | Prevalence of hip cartilage defects in non-OA participants |
| Chandra et al, 201636 | n=24; healthy asymptomatic volunteers (as per clinical examination, subjective scoring, and radiological evaluation (HOAMS).37 12M, 12F. Ages 23–34 years | 3T, single hip for each participant (side not specified), 3D fat-sat PD 3D-SPACE, multiecho spin-echo T2 map in the sagittal plane | Fully automated T2 assessment of femoral and acetabular cartilage vs manual segmentation. Automated method claimed to avoid measurement reliability/reproducibility issues | WJCV not explicitly stated; figure 5 demonstrates range of ∼4000–12 000 mm3 | HOAMS | None |
| Ramme et al, 201638 | n=20. Hip pain or hip OA. 5M, 15F. Ages 23–74 years | 3T, n=10 scanned with 3D true FISP, n=10 scanned with 3D GRE | Manual segmentations of the proximal femur and acetabular cartilage GRE and true FISP MR sequences, one observer considered gold standard. Third observer was automated rater. ICC for automated process vs gold standard rater was for GRE 0.286 and 0.614 for true FISP | WJCV estimated by an expert (8290–18 880 mm3), a physician (8420–21 330 mm3), and an automated computer algorithm (2300–9580 mm3) | None | None |
| Gallo et al, 201639 | n=54 participants. Longitudinal study on hip OA and FAI. FAI participants excluded. Exclusion criteria included knee OA (K-L>2) and hip K-L=4. 25F, 29M. Mean age 46.5±13.2 years | 3T, unilateral hip MR at baseline and 18 months for hip with greater K-L score. Semiquantitative: multiaxial intermediate-weighted fat-sat FSE. Quantitative: sagittal T1ρ/T2 mapping 3D SPGR. Fat-sat 3D MERGE | Two expert observers applied SHOMRI scoring system to evaluate the presence of articular cartilage lesions using the three FSE MRI series. SHOMRI has been shown to have high intra-rater and inter-rater reliability (ICC>0.91). Further detail in Lee et al 201440 | Semiquantitative assessment of cartilage defects as per SHOMRI.40 Hip OA progression per K-L scale. T1ρ and T2 relaxation times | K-L scale | At baseline, 16 participants (29.6%) had mild or moderate hip OA (K-L=2, 3). At 18-month follow-up, 9/54 (16.7%) demonstrated progression of hip OA. |
3D, three-dimensional; BML, bone marrow lesion; F, female; FAI, femoroacetabular impingement; fat-sat, fat-suppressed; FH, femoral head; FHCT, femoral head cartilage thickness; FHCV, femoral head cartilage volume; FSE, fast spin-echo; GRE, gradient-recalled echo; HOAMS, Hip Osteoarthritis MRI Scoring System; ICC, intraclass correlations; JSN, Joint Space Narrowing; K-L, Kellgren-Lawrence;41 M, male; MCCS, Melbourne Collaborative Cohort Study; MERGE, multiple echo recombined gradient echo; OA, osteoarthritis; PD, proton density; SHOMRI, scoring hip osteoarthritis with MRI; SPGR, segmented spoiled gradient-recalled acquisition; STIR, short T1 inversion recovery sequence; TASOACS, Tasmanian Older Adult Cohort; THR, total hip replacement; WJ, whole-joint; WJCT, whole-joint cartilage thickness; WJCV, whole-joint cartilage volume; WOMAC, Western Ontario and McMaster Universities Osteoarthritis index.42.