Reliability and Accuracy of Cross-sectional Radiographic Assessment of Severe Knee Osteoarthritis: Role of Training and Experience
Bibliographic record
Abstract
OBJECTIVE: To dêtermine the reliability of radiographic assessment of knee osteoarthritis (OA) by nonclinician readers compared to an experienced radiologist. METHODS: The radiologist trained 3 nonclinicians to evaluate radiographic characteristics of knee OA. The radiologist and nonclinicians read preoperative films of 36 patients prior to total knee replacement. Intrareader and interreader reliability were measured using the weighted κ statistic and intraclass correlation coefficient (ICC). Scores κ < 0.20 indicated slight agreement, 0.21-0.40 fair, 0.41-0.60 moderate, 0.61-0.80 substantial, and 0.81-1.0 almost perfect agreement. RESULTS: Intrareader reliability among nonclinicians (κ) ranged from 0.40 to 1.0 for individual radiographic features and 0.72 to 1.0 for Kellgren-Lawrence (KL) grade. ICC ranged from 0.89 to 0.98 for the Osteoarthritis Research Society International (OARSI) summary score. Interreader agreement among nonclinicians ranged from κ of 0.45 to 0.94 for individual features, and 0.66 to 0.97 for KL grade. ICC ranged from 0.87 to 0.96 for the OARSI Summary Score. Interreader reliability between nonclinicians and the radiologist ranged from κ of 0.56 to 0.85 for KL grade. ICC ranged from 0.79 to 0.88 for the OARSI Summary Score. CONCLUSION: Intrareader and interreader agreement was variable for individual radiograph features but substantial for summary KL grade and OARSI Summary Score. Investigators face tradeoffs between cost and reader experience. These data suggest that in settings where costs are constrained, trained nonclinicians may be suitable readers of radiographic knee OA, particularly if a summary score (KL grade or OARSI Score) is used to determine radiographic severity.
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 | 0.011 | 0.062 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| 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 | 0.000 |
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".