Variability in Interpretation of Magnetic Resonance Imaging of the Pediatric Sacroiliac Joint
Bibliographic record
Abstract
OBJECTIVE: Magnetic resonance imaging (MRI) is pivotal in the assessment of early sacroiliitis in children. We aimed to evaluate the agreement between local radiology reports and central imaging reviewers for active inflammation and structural damage at the sacroiliac (SI) joints. METHODS: Eight hospitals each contributed up to 20 cases of consecutively imaged children and adolescents with juvenile idiopathic arthritis and suspected sacroiliitis. Studies were independently reviewed by 3 experienced musculoskeletal pediatric radiologists. Local assessments of global impression and lesions were coded from the local radiology reports by 2 study team members. Test properties of local reports were calculated using the central imaging team's majority as the reference standard. RESULTS: For 120 evaluable subjects, the median age was 14 years, half of the cases were male, and median disease duration at the time of imaging was 0.8 years (interquartile range 0-2). Sensitivity of local reports for inflammation was high, 93.5% (95% confidence interval [95% CI] 78.6-99.2), and specificity was moderate, 69.7% (95% CI 59.0-79.0), but positive predictive value (PPV) was low, 51.8% (95% CI 38.0-65.3). Twenty-seven cases (23%) had active inflammation reported locally but rated normal at the central reading, 19 (70%) with subsequent medication changes. The sensitivity of local reports detecting structural damage was low, 45.7% (95% CI 28.8-63.4), and specificity was high, 88.2% (95% CI 79.4-94.2); PPV was low, 61.5% (95% CI 40.6-79.8). CONCLUSION: Substantial variation exists in the interpretation of inflammatory and structural lesions at the SI joints in children. To reliably identify pathology, additional training in the MRI appearance of the maturing SI joint is greatly needed.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.019 | 0.072 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".