Predicting cancer detection rates from multiparametric prostate MRI
Bibliographic record
Abstract
INTRODUCTION: Although the Prostate Imaging-Reporting and Data System (PI-RADS) categorization represents the standard method for assessing the risk of prostate cancer using prostate magnetic resonance imaging (MRI), there exists wide variation in cancer detection rates (CDRs) in real-world practice. We therefore evaluated the association of clinical and radiographic features with CDRs and developed a predictive model to improve clinical management. METHODS: We identified men aged 18-89 years with elevated prostate-specfic antigen (PSA) or on active surveillance for prostate cancer who underwent MRI-ultrasound (US) fusion biopsy or in-bore MRI-targeted biopsy. The associations of features with the per-lesion CDR (Gleason 6-10) and clinically significant (cs) CDR (Gleason 7-10) were examined using logistic regression, and results were operationalized into a predictive model. RESULTS: Targeted biopsy was performed for 347 lesions in 281 patients. Overall, the CDR was 49.0% and the csCDR was 28.0%. On multivariable analysis, increasing PI-RADS category, smaller prostate size, and increasing PSA density were independently associated with higher CDR, while prior prostate biopsy was associated with lower CDR. A solitary PI-RADS 3-5 lesion was independently associated with higher csCDR, while 2+ prior prostate biopsies was associated with a lower csCDR. A predictive model provided a greater net benefit than a strategy of performing biopsy in all PI-RADS 3-5 lesions across a wide range of threshold probabilities. CONCLUSIONS: Several clinical and radiographic features are independently associated with the risk of prostate cancer in men undergoing MRI-targeted biopsy. A predictive model based on these features can improve clinical decisions regarding biopsy compared to the conventional strategy of performing biopsy for all PI-RADS 3-5 lesions.
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 | 0.012 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| 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".