MA02.12 UC Screen California: A Statewide Participatory Informatics Approach to Lung Cancer Screening
Bibliographic record
Abstract
Canadian Early Detection of Lung Cancer, Pittsburgh Lung Screening Study, and Toronto Screening Program, we used Sybil to estimate lung cancer risk for year 1 to 6 and estimated the area under the receiving operating characteristics curve (AUC).We also evaluated the AUC by nodule diameter, solidity, and nodule presence.Results: Based on the 3 studies, the AUC for Sybil was 0.93 for lung cancer diagnosed within 1 year, down to 0.79 within 6 years (see Table 1).When restricting the analysis to those with pulmonary nodules, the AUCs of Sybil remained comparable.However, when focusing on individuals with no baseline nodules, the AUC from Sybil reduced to 0.65.When stratified by nodule diameter, Sybil performs better among larger nodules (maximum diameter > 10 mm), with AUCs of 0.91 for year 1 vs 0.86 in those with small nodules.When stratified by solidity, Sybil performs better in ground glass opacity nodules, with AUC above 0.93 in years 1 and 2, but it performs better in partially solid nodules in years 3 to 6, with AUC above 0.81.To assess if Sybil can predict lung cancer risk in the absence of nodules, we used LDCT scans for patients with a known lung cancer diagnosis but no apparent nodules at baseline.91 patients did not have nodules at baseline but were diagnosed with lung cancer between 1 to 6 years.83 of 91 had low Sybil scores at baseline, which suggests that Sybil's predictive performance may be suboptimal in the absence of nodules.Conclusions: We evaluated the predictive performance of Sybil in three independent lung cancer screening studies.The performance was best within 1 year.A decline in accuracy was observed for predictions after 3 years.We also found that Sybil may not be able to predict lung cancer risk without the presence of nodules and may have limited clinical use for fast-growing nodules, not present at baseline.
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 | 0.010 | 0.014 |
| Meta-epidemiology (narrow) | 0.000 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.002 | 0.003 |
| Science and technology studies | 0.002 | 0.000 |
| Scholarly communication | 0.002 | 0.000 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.011 | 0.002 |
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".