Purpose
This study evaluates the impact of second-opinion review of multiparametric prostate MRI for cancer detection by a multidisciplinary review board at a tertiary care center when compared with the initial community radiologist interpretation.
Methods
Cases were collected retrospectively from multidisciplinary prostate MRI rounds from 2017 to 2020 at a single tertiary care center. Patients with suspected prostate cancer or on active surveillance were referred for consideration of TRUS/MRI-fusion biopsy based on community-read prostate MRIs. All MRIs were re-read by subspecialized abdominal radiologists and a PI-RADS score assigned. Targeted fusion and 8–12 core systematic biopsy was performed in patients with PIRADS ≥ 3 lesions. Cohen kappa values were used to quantify interobserver agreement. Positive predictive value (PPV) was used to determine accuracy of PI-RADS score for detection of clinically significant prostate cancer (csPCa) (ISUP Grade Group ≥ 2).
Results
Three hundred and thirty-two lesions in 303 patients were reviewed and 252 lesions in 198 patients biopsied. The PI-RADS score was concordant in 60.5% of lesions, downgraded in 17.8%, and upgraded in 7.8%. Agreement between community and tertiary center interpretation was fair (
κ
= 0.354), with greater agreement for PI-RADS ≥ 4 (
κ
= 0.523) than PI-RADS ≥ 3 (
κ
= 0.456), and peripheral zone (
κ
= 0.419) than transition zone lesions (
κ
= 0.251). Prevalence of csPCa in biopsied lesions was 40.9%.
Conclusion
There is variability in community and tertiary care center interpretation of prostate MRI in cancer detection, with higher concordance rates for higher grade and peripheral zone lesions. These differences demonstrate the added value of multidisciplinary round review and highlight the need for ongoing education and feedback.
Graphical abstract
Purpose: To compare symbolic regression by genetic programming (SRGP) with symbolic regression by random search (SRRS), a novel method for symbolic regression described herein. Methods: We limit our problem space to N binary trees, m terminals and n functions, then use a dense enumeration of full binary trees to perform uniform random sampling from the set of all permitted equations. We compare a single basic configuration of symbolic regression by genetic programming with symbolic regression by random search using 1000 randomly generated problems. We perform a hyperparameter search with 50 randomly generated symbolic regression problems and 198 randomly generated hyperparameter configurations, examining the performance of SRGP against SRRS. Results: For the single configuration experiment, SRGP outperformed SRRS in 49.0% of problems, random search was best in 26.2% of problems, and there was a tie in 24.8% of problems. Of the cases that were not tied, genetic programming was best in 65.6% of experiments (99% CI, [60.7%, 69.2%]). Of the cases that were not tied in the hyperparameter search, SRGP was best in 44% (99% CI, [41%, 48%]) of cases. The average random configuration of SRGP performs worse than does SRRS. Conclusion: SRGP can outperform SRRS with appropriate hyperparameter selection, but our computations suggest that the average SRGP configuration performs worse than does SRRS.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.