Development, comparison, and validation of four intelligent, practical machine learning models for patients with prostate-specific antigen in the gray zone

Aug 4, 2026

Upload by Himbo

Machine learning prediction models are developed and compared for patients in the prostate-specific antigen (PSA) “gray zone” to identify valuable predictors and support real clinical decision-making. LogisticRegression, XGBoost, GaussianNB, and LGBMClassifier are trained using demographic, PSA-related parameters, prostate volume and PSA density features, plus prostate MRI results. Models are built after univariate and multivariate logistic analyses. Results show superior predictive performance versus single metrics, with LogisticRegression achieving the highest AUC and statistical advantage among compared models.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages11

Reading Time28 min

AccessFree to download

Purpose

Methods

Results

Conclusion

Introduction

TYPE Original Research PUBLISHED 08 June 2023 DOI 10.3389/fonc.2023.1157384 OPEN ACCESS EDITED BY Anthony Chi Fai Ng, The Chinese University of Hong Kong, Hong Kong SAR, China REVIEWED BY Albino Eccher, Integrated University Hospital Verona, Italy Ugo Giovanni Falagario, University of Foggia, Italy *CORRESPONDENCE Bin Fu [urofbin@163.com](urofbin@163.com)[ ](urofbin@163.com)Xinxi Deng [sudadengxinxi2011@163.com](sudadengxinxi2011@163.com)[ ](sudadengxinxi2011@163.com)Ru Chen [chenru99999@126.com](chenru99999@126.com) †These authors have contributed equally to this work RECEIVED 02 February 2023 ACCEPTED 24 May 2023 PUBLISHED 08 June 2023 CITATION Liu T, Zhang X, Chen R, Deng X and Fu B (2023) Development, comparison, and validation of four intelligent, practical machine learning models for patients with prostate-speciﬁc antigen in the gray zone. Front. Oncol. 13:1157384 . doi: 10.3389/fonc.2023.1157384 COPYRIGHT © 2023 Liu, Zhang, Chen, Deng and Fu. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Development, comparison, and validation of four intelligent, practical machine learning models for patients with prostate-speciﬁc antigen in the gray zone Taobin Liu 1,2†, Xiaoming Zhang 1†, Ru Chen 1*, Xinxi Deng 3* and Bin Fu 1,2* 1 Department of Urology, The First Afﬁliated Hospital of Nanchang University, Nanchang, China, 2Jiangxi Institute of Urology, Nanchang, Jiangxi, China, 3 Department of Urology, Jiu Jiang NO.1 People's Hospital, Jiujiang, China Purpose: Machine learning prediction models based on LogisticRegression, XGBoost, GaussianNB, and LGBMClassiﬁer for patients in the prostate-speciﬁc antigen gray zone are to be developed and compared, identifying valuable predictors. Predictive models are to be integrated into actual clinical decisions. Methods: Patient information was collected from December 01, 2014 to December 01, 2022 from the Department of Urology, The First Afﬁliated Hospital of Nanchang University. Patients with a pathological diagnosis of prostate hyperplasia or prostate cancer (any PCa) and having a prostatespeciﬁc antigen (PSA) level of 4–10 ng/mL before prostate puncture were included in the initial information collection. Eventually, 756 patients were selected. Age, total prostate-speciﬁc antigen (tPSA), free prostate-speciﬁc antigen (fPSA), fPSA/tPSA, prostate volume (PV), prostate-speciﬁc antigen density (PSAD), (fPSA/tPSA)/PSAD, and the prostate MRI results of these patients were recorded. After univariate and multivariate logistic analyses, statistically signiﬁcant predictors were screened to build and compare machine learning models based on LogisticRegression, XGBoost, GaussianNB, and LGBMClassiﬁer to determine more valuable predictors. Results: Machine learning prediction models based on LogisticRegression, XGBoost, GaussianNB, and LGBMClassiﬁer exhibit higher predictive power than individual metrics. The area under the curve (AUC) (95% CI), accuracy, sensitivity, speciﬁcity, positive predictive value, negative predictive value, and F1 score of the LogisticRegression machine learning prediction model were 0 . 932 (0 .881– 0.983), 0.792, 0.824, 0.919, 0.652, 0.920, and 0.728, respectively; of the XGBoost machine learning prediction model were 0.813 (0.723–0.904), 0.771, 0.800, 0.768, 0.737, 0.793 and 0.767, respectively; of the GaussianNB machine learning prediction model were 0.902 (0.843–0.962), 0.813, 0.875, 0.819, 0.600, 0.909, and 0.712, respectively; and of the LGBMClassiﬁer machine learning prediction model were 0.886 (0.809–0.963), 0.833,

FAQs

What patient group and PSA range define the study population?

Patients are selected with prostate hyperplasia or prostate cancer and a PSA level of 4–10 ng/mL before prostate puncture, corresponding to the PSA gray zone.

Which machine learning models are developed and compared?

The study builds and compares LogisticRegression, XGBoost, GaussianNB, and LGBMClassifier models.

What key outcome shows LogisticRegression performs best?

LogisticRegression achieves the highest AUC, and the AUC difference compared with XGBoost, GaussianNB, and LGBMClassiﬁer is statistically significant (p < 0.001).

Aug 4, 2026

Upload by Himbo