Article on Prostate Cancer Predictive Model and Risk Calculator
International Team Develops AI-Based Tool to Improve Prostate Cancer Diagnosis
A multinational team of urologists and data scientists has developed a novel machine learning-based predictive model to enhance the detection of clinically significant prostate cancer (csPCa) in men undergoing prostate biopsy. Recently published in Urology, the study introduces an innovative online risk calculator that provides clinicians with a fast and accessible method to estimate an individual patient’s likelihood of harboring csPCa using standard clinical variables.
The model was trained and externally validated on retrospective data from over 1,200 patients who underwent transperineal prostate biopsies across three countries: Australia, New Zealand, and Switzerland. Leveraging the LightGBM algorithm, the model incorporates key predictors such as age, PSA density, and multiparametric MRI findings.
Unlike traditional risk calculators, which are typically based on logistic regression and often developed within single-country populations, this AI-driven model demonstrated superior discrimination (as measured by AUC), robust calibration, and high clinical utility based on decision curve analysis. Its strong performance across both internal and external validation cohorts – including those who underwent transrectal ultrasound-guided (TRUS) biopsies – supports its generalizability and potential for broad clinical adoption.
To facilitate real-world use, the team created a free, web-based risk calculator tailored for use during urology consultations. Clinicians can input patient-specific data to obtain a real-time estimate of csPCa risk, thereby enhancing shared decision-making and helping reduce unnecessary biopsies and overtreatment.
‘Our goal was to create a clinically relevant, user-friendly tool that reflects the heterogeneity of real-world patients and improves risk stratification. This model can support more precise decision-making and help urologists focus biopsy resources on those who truly need them.’ said first author Dr. Flavio Ordones, a consultant uro-oncologist based in New Zealand.
The authors emphasize the need for future prospective studies to further evaluate the tool’s clinical impact, particularly in reducing biopsy rates and guiding treatment decisions. Nonetheless, these findings mark a promising advancement in the integration of artificial intelligence into prostate cancer diagnostics.
The online calculator is now publicly accessible and designed for seamless integration into global urological practice.
Title: A Novel Machine Learning-based Predictive Model of Clinically Significant Prostate Cancer and Online Risk Calculator
Authors: Flavio Vasconcelos Ordones, Paulo Roberto Kawano, Lodewikus Vermeulen, Ali Hooshyari, David Scholtz, Peter John Gilling, Darren Foreman, Basil Kaufmann, Cedric Poyet, Michael Gorin, Abner Macola Pacheco Barbosa, Naila Camila da Rocha, Luis Gustavo Modelli de Andrade.
You can read the Full Article on Urology.
Find more posts about Prostate Cancer on OncoDaily.
-
Challenging the Status Quo in Colorectal Cancer 2024
December 6-8, 2024
-
ESMO 2024 Congress
September 13-17, 2024
-
ASCO Annual Meeting
May 30 - June 4, 2024
-
Yvonne Award 2024
May 31, 2024
-
OncoThon 2024, Online
Feb. 15, 2024
-
Global Summit on War & Cancer 2023, Online
Dec. 14-16, 2023