Evaluation of Automated Fracture Risk Assessment Based on the Canadian Association of Radiologists and Osteoporosis Canada Assessment Tool
Bibliographic record
Abstract
Fracture risk assessments are not always clearly communicated on bone mineral density (BMD) reports; evidence suggests that structured reporting (SR) tools may improve report clarity. The aim of this study is to compare fracture risk assessments automatically assigned by SR software in accordance with Canadian Association of Radiologists and Osteoporosis Canada (CAROC) recommendations to assessments from experts on narrative BMD reports. Charts for 500 adult patients who recently received a BMD exam were sampled from across University of Toronto's Joint Department of Medical Imaging. BMD measures and clinical details were manually abstracted from charts and were used to create structured reports with assessments generated by a software implementation of CAROC recommendations. CAROC calculations were statistically compared to experts' original assessments using percentage agreement (PA) and Krippendorff's alpha. Canadian FRAX calculations were also compared to experts', where possible. A total of 25 (5.0%) reported assessments did not conform to categorizations recommended by Canadian guidelines. Across the remainder, the Krippendorff's alpha relating software assigned assessments to physicians was high at 0.918; PA was 94.3%. Lower agreement was associated with reports for patients with documented modifying factors (alpha = 0.860, PA = 90.2%). Similar patterns of agreement related expert assessments to FRAX calculations, although statistics of agreement were lower. Categories of disagreement were defined by (1) gray areas in current guidelines, (2) margins of assessment categorizations, (3) dictation/transcription errors, (4) patients on low doses of steroids, and (5) ambiguous documentation of modifying factors. Results suggest that SR software can produce fracture risk assessments that agree with experts on most routine, adult BMD exams. Results also highlight situations where experts tend to diverge from guidelines and illustrate the potential for SR software to (1) reduce variability in, (2) ameliorate errors in, and (3) improve clarity of routine adult BMD exam reports.
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 | 0.004 | 0.019 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.005 | 0.002 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.003 | 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".