Radiologist Interpretive Volume and Breast Cancer Screening Accuracy in a Canadian Organized Screening Program
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
BACKGROUND: To strengthen evidence on which radiologist mammography interpretive volume requirements can be based, we assessed the relation of volume to accuracy in the Quebec Breast Cancer Screening Program. METHODS: Annual interpretive volume (total, screening, and diagnostic) for all 340 radiologists who interpreted 1315327 screening examinations in the period from 2000 to 2006 was obtained using provincial databases. The association of volume to sensitivity, false-positive rate, and accuracy (sensitivity/false-positive rate) was assessed by multivariable Poisson regression with robust error variance. All statistical tests were two-sided. RESULTS: Radiologists consistently interpreting less than 500 mammograms annually experienced a 58% reduction in accuracy (adjusted accuracy ratio = 0.42; 95% confidence interval [CI] = 0.24 to 0.74) compared with those who consistently interpreted at least 500 mammograms annually. Moreover, accuracy increased progressively as total annual volume increased (P trend = .0005). Radiologists interpreting at least 4000 mammograms annually experienced a 32% increase in accuracy (adjusted accuracy ratio = 1.32; 95% CI = 1.13 to 1.54) compared with those interpreting 500 to 999 mammograms annually. This increase in accuracy is attributable to a reduction in false-positive rate as total volume increased (P trend = .001). Sensitivity changed little with total volume (P trend = .68). Gains in accuracy were greater up to approximately 3000 mammograms interpreted annually. CONCLUSIONS: The minimum annual volume of 500 mammograms required in North America is justified; radiologist accuracy may be compromised if interpretive volume is consistently less than this requirement. Raising interpretive volume may help to reduce the frequency of false positives without loss of sensitivity. Possible gains in accuracy may be greater with increases in volume of up to approximately 3000 mammograms interpreted annually.
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 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 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 | 0.000 |
| 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