Machine Learning-Based Models Enhance the Prediction of Prostate Cancer

Abstract

PurposePSA is currently the most commonly used screening indicator for prostate cancer. However, it has limited specificity for the diagnosis of prostate cancer. We aim to construct machine learning-based models and enhance the prediction of prostate cancer.MethodsThe data of 551 patients who underwent prostate biopsy were retrospectively retrieved and divided into training and test datasets in a 3:1 ratio. We constructed five PCa prediction models with four supervised machine learning algorithms, including tPSA univariate logistic regression (LR), multivariate LR, decision tree (DT), random forest (RF), and support vector machine (SVM). The five prediction models were compared based on model performance metrics, such as the area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, specificity, calibration curve, and clinical decision curve analysis (DCA).ResultsAll five models had good calibration in the training dataset. In the training dataset, the RF, DT, and multivariate LR models showed better discrimination, with AUCs of 1.0, 0.922 and 0.91, respectively, than the tPSA univariate LR and SVM models. In the test dataset, the multivariate LR model exhibited the best discrimination (AUC=0.918). The multivariate LR model and SVM model had better extrapolation and generalizability, with little change in performance between the training and test datasets. Compared with the DCA curves of the tPSA LR model, the other four models exhibited better net clinical benefits.ConclusionThe results of the current retrospective study suggest that machine learning techniques can predict prostate cancer with significantly better AUC, accuracy, and net clinical benefits.

Unleashing the potential: Artificial intelligence in urology for enhanced diagnosis, treatment, and personalized care

“…Incorporating multiple modalities from different sources such as risk factors, imaging, and pathology reading can maximize the accuracy of prediction and diagnosis. Not only biomarkers such as prostate-specific antigen and biopsy results but also MRI imaging data can be used to improve clinical accuracy and efficiency in PCa diagnosis [ 2 3 ]. Also, Ström et al [ 4 ] developed an image analysis tool for the diagnosis and grading of prostate cancer based on histopathological slides.…”

ProsTAV, a novel blood‐based test for biopsy decision management in significant prostate cancer

“…30 Machine learning has emerged as a powerful tool to derive predictive models for PCa screening or prognosis. 13,31 In this study, we combined analysis of TAV with machine learning to implement a predictive machine learning model, with the ultimate goal of reducing the number of unnecessary biopsies during early diagnosis of PCa. and Houston Methodist, US) between 2019 and 2021.…”

“…During machine learning, computers are programmed to identify patterns from the data, which can be very advantageous when analyzing large datasets 30 . Machine learning has emerged as a powerful tool to derive predictive models for PCa screening or prognosis 13,31 …”

Knowledge structure and emerging trends of AR variants in prostate cancer: a bibliometric analysis based on CiteSpace and VOSviewer

“…Prostate cancer (PCa) is a common malignant cancer among elderly men. PCa incidence increases yearly with the aging of population, which results in heavy disease burden [1]. PCa mortality rate is decreasing with the therapies development however it still poses threat to the health of middle-aged and elderly men [2].…”