Identification of Factors Affecting Prostate Cancer Using Machine Learning Methods: A Systematic Review

Area of Science:

• Oncology
• Data Science
• Bioinformatics

Background:

• Prostate cancer is a leading global malignancy, necessitating effective prevention and screening.
• Identifying risk factors is crucial for managing the rising incidence and mortality rates.
• Artificial intelligence and machine learning offer advanced tools for risk factor analysis.

Purpose of the Study:

• To systematically review machine learning applications in identifying prostate cancer risk factors.
• To determine the most frequently used machine learning methods and data sources in this field.
• To guide future research for enhanced prostate cancer prevention and screening.

Main Methods:

• Systematic review of articles from PubMed, Scopus, Web of Science, and IEEE Xplore (2015-2024).
• Adherence to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.
• Inclusion criteria focused on studies using machine learning to investigate prostate cancer risk factors.

Main Results:

• China leads research in machine learning for prostate cancer risk factors.
• Key identified risk factors include age, prostate-specific antigen (PSA), total PSA (tPSA), free PSA (fPSA), and PSA density (PSAD).
• Random forest, support vector machine, and logistic regression were prevalent ML methods; R and Python were common analysis tools. SEER, PLCO, and NCBI were frequently used data sources.

Conclusions:

• Machine learning provides powerful tools for identifying significant prostate cancer risk factors.
• Utilizing registered data sources and advanced ML methods can improve risk prediction.
• Findings support enhanced strategies for prostate cancer prevention and early detection.