Limited Reliability of Radiographic Assessment of Sacroiliac Joints in Patients with Suspected Early Spondyloarthritis
Bibliographic record
Abstract
OBJECTIVE: To determine the reproducibility of evaluation of sacroiliac joint (SIJ) radiographs among readers with varying levels of experience, and to identify potential drivers of disagreement in classification among 5 predefined radiographic lesion types. METHODS: The study sample consisted of 104 consecutive patients aged 18-40 with low back pain ≥ 3 months of duration who met the Assessment of SpondyloArthritis international Society (ASAS) definition for a positive SIJ magnetic resonance image, or were HLA-B27-positive and had ≥ 1 spondyloarthritis (SpA)-related clinical/laboratory feature according to the ASAS classification criteria for axial SpA. Seven blinded readers (2 musculoskeletal radiologists, 5 rheumatologists) classified pelvic radiographs according to the modified New York criteria (mNY) and recorded presence/absence of 5 lesion types in both SIJ: erosion, sclerosis, ankylosis, joint space widening, and joint space narrowing. Reproducibility of mNY classification among 21 reader pairs was assessed and potential drivers of disagreement were identified among 5 lesion types. A generalized linear mixed logistic regression model served to analyze to what extent discordance in lesion type was associated with discrepant mNY classification. RESULTS: Mean κ values (percent concordance) were 0.39 (84.1%) for mNY classification over 21 reader pairs, 0.46 (79.8%) between 2 musculoskeletal radiologists, and 0.55 (86.5%) and 0.36 (77.9%) between the most experienced rheumatologist and the 2 radiologists. Erosion showed the lowest agreement (25%) among patients with discordant classification and gave the highest OR of 13.5 for disagreement. CONCLUSION: Reproducibility of radiographic SIJ classification in an SpA inception cohort was only fair to at best moderate among 7 readers with varying levels of experience, questioning the applicability of mNY in early SpA.
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 | 0.017 | 0.094 |
| Meta-epidemiology (narrow) | 0.000 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 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 | 0.001 |
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".