PA 17-4-2588 Estimating historical exposure without imputation: lessons learned from a predictive modeling approach using data from a cohort of ontario uranium underground miners in canada
Bibliographic record
Abstract
Uranium underground miners are exposed to a number of radionuclides that undergo radioactive decay. Historically, studies examining adverse health effects have been focused largely on alpha radiation (radon gas). However, recent international studies have shown evidence of excess mortality of lung cancer and leukemia with increased cumulative doses of external gamma radiation. The objective of this study is to develop and validate a predictive model for estimating gamma radiation exposure for miners working in uranium mines and to apply this exposure information to derive health-based risks. The dose prediction model was developed and validated using a cross-validation approach. To aid in model development, 70% random sample of workers were used in the model development (i.e., Training Sample) while the remainder 30% (i.e., Test Sample) was used to determine model performance. ROBUSTREG in SAS was used to minimise the effects of outliers. Poisson regression was used to derive relative-risks (RR). Regression analysis showed that individual dosimetric readings were modestly predicted by individual work history and geological characteristics of Ontario uranium mines (p<0.001, R2=0.374). Preliminary risk estimates were conducted for a subset of the OUM cohort as proof-of-concept for the reconstruction of historical gamma exposure. In total, there were 12 953 miners that contributed 4 31 655 person-years of observation from 1954 to 1992. There was a non-significant increase in lung cancer mortality (RR=1.11, 95% CI: 0.85 to 1.45), and a significant increased risk of all forms of leukemia, when comparing the highest cumulative dose category (>14 mSievert (mSv)) to the reference category (0 mSv) (RR=2.58, 95% CI: 1.06 to 6.30). When measured exposure data is not available, predictive modelling can be an effective way to estimate historical exposure without imputation that in turn used to derive health-based risk estimates.
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 | 0.007 | 0.023 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| 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".