# Convolutional neural networks for prostate cancer detection, classification, and segmentation: A systematic review and bibliometric analysis

## TL;DR

This paper reviews how convolutional neural networks are being used to detect, classify, and segment prostate cancer in medical imaging, highlighting progress and challenges in the field.

## Contribution

The study provides a systematic review and bibliometric analysis of CNN applications in prostate cancer, identifying trends and gaps in research.

## Key findings

- Vision Transformer models achieved the highest classification accuracy among evaluated CNN architectures.
- U-Net variants were most commonly used for segmentation with high Dice coefficients.
- SHAP was the most frequently adopted explainability method for clinical interpretability.

## Abstract

Prostate cancer represents the second most common malignancy among men globally, necessitating accurate diagnostic methodologies for optimal patient outcomes. Convolutional neural networks (CNNs), a core deep learning methodology, have emerged as transformative technologies for automated prostate cancer detection, classification, and segmentation across multiple imaging modalities.

A systematic review following PRISMA guidelines was conducted across Web of Science, Scopus, and PubMed databases (January 2020–December 2025). CNN-based classification architectures were analyzed across ResNet, Vision Transformer, DenseNet, Xception, ConvNeXT, and Swin Transformer implementations, with comparative evaluation of accuracy and transfer learning performance. Object detection and segmentation approaches were examined across U-Net variants, R-CNN family algorithms, and YOLO-based implementations. Hyperparameter optimization strategies were assessed. Explainable AI methodologies including SHAP, Grad-CAM, DiCE, and LIME were evaluated for clinical interpretability and spatial localization accuracy.

Analysis of 320 publications revealed peak research activity in 2024 (63 publications, 19.7%). The United States led with 58 publications (18.1%), followed by China with 55 (17.2%). Multiparametric MRI constituted the primary imaging modality (42.5%), followed by histopathology (28.1%), ultrasound (14.1%), and PET imaging (9.4%). Vision Transformer models demonstrated the highest classification accuracy among evaluated architectures, while U-Net variants dominated segmentation applications with consistently high Dice coefficients. SHAP emerged as the most frequently adopted explainability method across the reviewed studies.

CNN-based prostate cancer detection, classification, and segmentation demonstrate promise for improving diagnostic accuracy and clinical workflow efficiency, though challenges in dataset standardization, regulatory compliance, and clinical integration remain to be addressed.

•Comprehensive analysis of publications reveals CNN advancement in prostate cancer detection.•Vision Transformers achieve high accuracy while U-Net dominates segmentation applications.•SHAP explainability methods appear in huge ratio of studies addressing clinical interpretability.•Dataset heterogeneity and computational constraints limit widespread clinical translation.•Multi-modal integration and federated learning present promising future research directions.

Comprehensive analysis of publications reveals CNN advancement in prostate cancer detection.

Vision Transformers achieve high accuracy while U-Net dominates segmentation applications.

SHAP explainability methods appear in huge ratio of studies addressing clinical interpretability.

Dataset heterogeneity and computational constraints limit widespread clinical translation.

Multi-modal integration and federated learning present promising future research directions.