A quantitative image analysis model of prostate biopsies for predicting clinical risk in men enrolled in an active surveillance program.
Bibliographic record
Abstract
111 Background: Quantitative image analysis of the prostate needle biopsy (PNB) has proven to be a robust and predictive platform for prostate cancer (PCa) prognosis. We sought to determine the performance of quantitative metrics in identifying which patients enrolled in an active surveillance (AS) program are most likely to present with significant clinical risk, including subsequent biopsy Gleason grade (GG) upgrading and/or a short (less than 24 month) prostate-specific antigen (PSA) doubling time (PSADT). Methods: One hundred sixty two AS patients (median age 70, 94% cT1-T2a, 85% <=GS6, median PSA 5.9 ng/mL) with overall 8 year median follow up and available diagnostic PNB specimens were analyzed. Computerized image analysis derived quantitative biometric features representing PCa morphology and immunofluorescent (IF) biomarkers from the PNB. Multivariate models predicting either GG upgrading on a subsequent PNB or a PSADT less than 24 months were evaluated. The AUC/concordance index (CI), sensitivity, specificity and hazard ratio (HR) were used to assess performance. Results: Univariate distribution of selective features, notably expression levels of AR and Ki67, were reflective of a low risk cohort.A multivariate model with three quantitative imaging metrics was trained with a CI of 0.77. Men at high risk within 24 months of the PNB were identified with 80% sensitivity and 73% specificity, HR of 5.7. The most important feature measured the relative proportion of tumor epithelial nuclei that were both Androgen Receptor and alpha-methylacyl-CoA racemase positive. The other two features were morphological assessments of epithelial cellular area compared to luminal area. Of note, clinical features such as age, GG and PSA were not selected in competition with the imaging metrics. Conclusions: Quantitative image analysis of morphology and IF biomarker expression in the PNB outperformed standard clinical features in a multivariate model to accurately predict which AS patients are at risk for Gleason upgrading and/or a shortened PSADT. Identifying such patients may prove beneficial in the primary treatment decision process.
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 | 0.002 | 0.004 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 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".