This study is a cross-sectional observational study conducted between January 2024 and June 2024. The study included individuals diagnosed with ERA who were presented with sacroiliitis. Diagnosis of ERA was based on the classification criteria established by the International League of Associations for Rheumatology (ILAR) [1]. The required minimum sample size was calculated using priori-power analysis. The priori-power analysis with α (significance level) = 0.05 and 1–β (power) = 0.90 was conducted using PASS (Power Analysis and Sample Size) version 14.0 software to determine the required minimum sample size for an intraclass correlation coefficient (ICC) above 0.95 (almost perfect agreement). Since there was no similar study available in the literature to use as a reference, and no opportunity to conduct a pilot study, the priori-power analysis was performed to achieve an ICC value greater than 0.95, which corresponds to an “almost perfect agreement.” The minimum sample size was calculated as 14. To improve the power of our study, we decided to study with 20 patients. We also performed a post-hoc power analysis using our obtained results to determine the power of our study. The post-hoc power analysis was conducted using the ICC between CT and ZTE obtained for total score (ICC = 0.99), sample size n = 20, and significance level (α) = 0.05. The power of our study was calculated to be 96%.

The study meticulously documented demographic data, clinical manifestations, and laboratory findings (including white blood cell [WBC] count, erythrocyte sedimentation rate [ESR], C–reactive protein [CRP], and HLA-B27). It also included a comprehensive examination of treatment modalities.

The exclusion criteria comprised patients with psoriasis, familial Mediterranean fever, and inflammatory bowel disease.

Disease activity was assessed using the Juvenile Spondyloarthritis Disease Activity Index (JSpADA), a validated tool designed to measure disease activity specifically in juvenile spondyloarthritis [5].

The study adhered to the Helsinki Declaration guidelines for medical research involving human subjects and received approval from the local ethics committee. Written informed consents were obtained from patients and their parents.

All MRI procedures were performed by using a 1.5-T (T) scanner (GE Healthcare, Milwaukee, WI) with a 16-channel air coil. The MRI protocol includes a fat-suppressed axial T2-weighted sequence (TR/TE 2,500/85 ms, echo train length 16), axial T1-weighted sequence (TR/TE 610/minimum ms, echo train length 3), coronal T1-weighted sequence (TR/TE 590/minimum ms, echo train length 3), coronal short tau inversion recovery (STIR) sequence (TR/TE 4,250/42 ms, echo train length 16), and axial T2-weighted sequence (TR/TE 4,650/85 ms, echo train length 16). The imaging parameters included a 4-mm section thickness, a 1-mm intersection gap, a 384 × 384 matrix, and a 26 cm × 26 cm field of view. Additionally, a ZTE sequence (TR/TE 568/0 ms, 1.5-mm section thickness, no intersection gap, a 280 × 280 matrix, and a 40 cm × 40 cm field of view, flip angle 2°) was performed.

Active sacroiliitis was assessed by MRI according to the Assessment of Spondyloarthritis International Society (ASAS) criteria [6].

To minimize temporal changes, low-dose CT and MRI were performed within 1 week. The CT scan was conducted using a 640-MSCT device, exposing patients to an average effective dose of 0.615 mSv per sacroiliac CT (Canon Medical Systems, Otowara, Japan). The scanning parameters included a voltage of 120 kV, automatically determined current based on the patient’s volume (80–95 mAs), a rotation time of 0.5 s, a layer thickness of 0.5 mm, a reconstruction interval of 2.5 mm, a window width of 2,700 Hounsfield units (HU), and a window level of 350 HU. No intravenous contrast agent was used.

All MRI and CT scans were independently and anonymously evaluated by two pediatric radiologists, T.O., with 10 years of experience, and Y.A., with 25 years of experience, who were blinded to the patients’ information. All scores were conducted by two independent observers. ICC was used to determine the agreement between them. All imaging studies were independently evaluated by both readers following the same sequence and order. The assessments were conducted in the following order: first, conventional MRI; then, ZTE; and finally, low-dose CT. All images were anonymized and reviewed twice by each reader at separate time points, with the patient order randomized for each evaluation. A 2-day interval was maintained between the first and second assessments for each modality to minimize the likelihood of readers recalling the images or scores of the same patient. Structural lesions (erosion, sclerosis, joint space changes) were scored in the coronal plane. The joint was divided into anterior, middle, and posterior thirds [7]. From each third, the sections with the most intense structural changes were selected, ensuring that the selected sections were not consecutive or adjacent. While our scoring was based on a single section, we conducted a comprehensive evaluation of all patients and all sequences. Sclerosis was defined as the presence of sclerosis exceeding 5 mm in the subchondral area.

All statistical analyses were performed using IBM SPSS for Windows version 29.0 (IBM Corp., Armonk, New York, NY). Shapiro–Wilk’s test was used to assess the normality assumption. Since the normality assumption did not hold, continuous variables were presented with median and interquartile range (IQR). Categorical variables were presented with the number of observations and percentages. Friedman’s two-way analysis of variance (ANOVA) was conducted for dependent group comparisons. Dunn’s test was used for the pairwise multiple comparisons. Using low-dose CT as the reference standard, the agreement between ZTE and low-dose CT was determined by ICC. These parameters were assessed at the quadrant level for erosion and sclerosis, and at the joint level for joint space changes (Supplementary Material 1). A P-value < 0.05 was considered statistically significant.