Exclusion of Focal Vertebral Artifacts From Spine Bone Densitometry and Fracture Prediction: A Comparison of Expert Physicians, Three Computer Algorithms, and the Minimum Vertebra
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
UNLABELLED: Expert physicians and automated methods for the exclusion of vertebral levels in DXA scans containing focal artifacts were compared. All methods of vertebral exclusion led to a small improvement in fracture prediction. Computer algorithms performed at least as well as physicians. INTRODUCTION: Lumbar spine DXA is often confounded by focal artifacts. Clinical rules and automated methods for vertebral exclusion have been proposed, but their concordance, effect on diagnosis, and fracture prediction is unknown. MATERIALS AND METHODS: We analyzed clinical DXA scans of the lumbar spine (20,478 women and 1534 men) performed from 1998 to 2002 (Province of Manitoba, Canada). Longitudinal health service records were assessed for the presence of nontrauma fracture codes after BMD testing. The effect of vertebral exclusions by expert physicians and several automated methods on diagnosis and prediction of incident fractures was compared. RESULTS: Vertebral exclusions were reported by physicians in over one quarter of the scans (31% of women and 29% of men). All methods of vertebral exclusion significantly decreased the mean spine T-score and increased the proportion of women designated as osteoporotic. kappa values and ROC area under the curve (AUC) for physician-computer agreement in the identification of abnormal scans indicated fair to moderate agreement in both women and men. Compared with no vertebral exclusions, a small increase in the hazard ratio and AUC for spine fracture and osteoporotic fracture prediction was seen after physician and computer exclusions. Compared with physician exclusions, AUC for prediction of osteoporotic fractures in men increased significantly with one computer algorithm (p = 0.004). The minimum vertebral T-score enhanced fracture prediction compared with no exclusions but approximately doubled the prevalence of osteoporotic categorization. CONCLUSIONS: We observed fair to moderate agreement between the physician and computer methods for vertebral level exclusion. All methods of vertebral exclusion led to a small improvement in fracture prediction using the lumbar spine measurement. The automated algorithms performed at least as well as physicians when fractures were used as the endpoint.
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.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 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 it