Abstract
Precise Gleason grading of prostate biopsy specimens is vital for determining the appropriate clinical management of prostate cancer. However, traditionally, subjective manual evaluation by pathologists is susceptible to inter-observer variability, contributing to variable diagnoses and a likelihood of less-than-optimal treatment decisions. Therefore, we present a hybrid deep-learning architecture, wherein a modified ResNet50 convolutional backbone has been amalgamated with a Vision Transformer (ViT) module with the aim of automated and standardized Gleason classification. The ResNet50 portion consists of 50 layers with bottleneck residual blocks inserted for texture and glandular pattern localization in contrast-enhanced histopathological images. The spatially rich feature maps are then forwarded to the ViT module that extracts long-range dependencies and contextual relationships across image patches through a combination of multi-head self-attention mechanisms and transformer encoders. In this manner, a combination of local feature extraction and global attention facilitates the model's learning of subtle morphological variations that are crucial for the differentiation of six different Gleason patterns on a large scale. The model was trained and validated on a balanced multiclass dataset of prostate biopsy images, achieving a classification accuracy of 99%, which is better than several existing deep-learning baselines. This hybrid architecture aims to enhance diagnostic consistency while providing a realistic, interpretable framework for implementation in clinical workflows geared toward high-throughput prostate cancer screening, especially in resource-limited healthcare settings.
References
Egevad, Lars, Chiara Micoli, Brett Delahunt, Hemamali Samaratunga, Hans Garmo, Pär Stattin, and Martin Eklund. "Gleason scores provide more accurate prognostic information than grade groups." Pathology (2025).
Sarıateş, Murat, and Erdal Özbay. "A Classifier Model Using Fine-Tuned Convolutional Neural Network and Transfer Learning Approaches for Prostate Cancer Detection." Applied Sciences 15, no. 1 (2024): 225.
Sungur, Mustafa, Aykut Aykaç, Mehmet Erhan Aydin, Ozer Celik, and Coskun Kaya. "Machine Learning-Based Prediction of Prostate Biopsy Necessity Using PSA, MRI, and Hematologic Parameters." Journal of Clinical Medicine 14, no. 1 (2024): 183.
Bottillo, Irene, Alessandro Sciarra, Giulio Bevilacqua, Alessandro Gentilucci, Beatrice Sciarra, Valerio Santarelli, Stefano Salciccia et al. "Early Detection of the Pathogenetic Variants of Homologous Recombination Repair Genes in Prostate Cancer: Critical Analysis and Experimental Design." Biology 14, no. 2 (2025): 117.
Li, Gaojie, Ye Wang, Ying Wang, Baojun Wang, Yuan Liang, Ping Wang, Yudan He et al. "PCaseek: ultraspecific urinary tumor DNA detection using deep learning for prostate cancer diagnosis and Gleason grading." Cell Discovery 10, no. 1 (2024): 90.
Kong, Fei, Xiyue Wang, Jinxi Xiang, Sen Yang, Xinran Wang, Meng Yue, Jun Zhang et al. "Federated attention consistent learning models for prostate cancer diagnosis and Gleason grading." Computational and Structural Biotechnology Journal 23 (2024): 1439.
Harder, Christian, Alexey Pryalukhin, Alexander Quaas, Marie-Lisa Eich, Maria Tretiakova, Sebastian Klein, Alexander Seper et al. "Enhancing Prostate Cancer Diagnosis: Artificial Intelligence–Driven Virtual Biopsy for Optimal Magnetic Resonance Imaging-Targeted Biopsy Approach and Gleason Grading Strategy." Modern Pathology 37, no. 10 (2024): 100564.
Alici-Karaca, Demet, and Bahriye Akay. "An efficient deep learning model for prostate cancer diagnosis." IEEE Access (2024).
Patel, Maulika, Parag Sanghani, and Niraj Shah. "Prostate Cancer Gleason Grading: A Review on Deep Learning Approaches for Recognizing." In ITM Web of Conferences, vol. 65, p. 03013. EDP Sciences, 2024
Butt, Muhammad Asim, Muhammad Farhat Kaleem, Muhammad Bilal, and Muhammad Shehzad Hanif. "Using multi-label ensemble CNN classifiers to mitigate labelling inconsistencies in patch-level Gleason grading." Plos one 19, no. 7 (2024): e0304847.
Sethi, Bijaya Kumar, Debabrata Singh, Saroja Kumar Rout, and Sandeep Kumar Panda. "Long short-term memory-deep belief network-based gene expression data analysis for prostate cancer detection and classification." IEEE Access 12 (2023): 1508-1524.
Varan, Metin, Jahongir Azimjonov, and Bilgen Maçal. "Enhancing prostate cancer classification by leveraging key radiomics features and using the fine-tuned linear svm algorithm." Ieee Access 11 (2023): 88025-88039.
Tolkach, Yuri, Vlado Ovtcharov, Alexey Pryalukhin, Marie-Lisa Eich, Nadine Therese Gaisa, Martin Braun, Abdukhamid Radzhabov et al. "An international multi-institutional validation study of the algorithm for prostate cancer detection and Gleason grading." NPJ precision oncology 7, no. 1 (2023): 77.
W. Bulten et al., “Artificial intelligence for diagnosis and Gleason grading of prostate cancer: the PANDA challenge,” Nature Medicine, vol. 28, no. 1, 2022, 154–163. https://doi.org/10.1038/s41591-021-01620-2.
Li, Bochong, Ryo Oka, Ping Xuan, Yuichiro Yoshimura, and Toshiya Nakaguchi. "Robust multi-modal prostate cancer classification via feature autoencoder and dual attention." Informatics in Medicine Unlocked 30 (2022): 100923.
Malibari, Areej A., Reem Alshahrani, Fahd N. Al-Wesabi, SB Haj Hassine, Mimouna Abdullah Alkhonaini, and Anwer Mustafa Hilal. "Artificial intelligence based prostate cancer classification model using biomedical images." Comput. Mater. Contin 72 (2022): 3799-3813.
Liu, Bojing, Yinxi Wang, Philippe Weitz, Johan Lindberg, Johan Hartman, Wanzhong Wang, Lars Egevad, Henrik Grönberg, Martin Eklund, and Mattias Rantalainen. "Using deep learning to detect patients at risk for prostate cancer despite benign biopsies." Iscience 25, no. 7 (2022).
Hammouda, Kamal, Fahmi Khalifa, Moumen El-Melegy, Mohamed Ghazal, Hanan E. Darwish, Mohamed Abou El-Ghar, and Ayman El-Baz. "A deep learning pipeline for grade groups classification using digitized prostate biopsy specimens." Sensors 21, no. 20 (2021): 6708.
M., Kusuma Sri, and Sathees Kumaran S. "Quadratic Luminance Vision Transformer Attention Network for Automated Mitosis Detection in Breast Histopathology Images." Journal of Innovative Image Processing 7, no. 2 (2025): 266-289
Tyagi, Nandini, Riya Sharma, and Monika Sharma. "Deep Learning for Skin Cancer Classification: A Study of Model Accuracy, Generalization, and Ensemble Learning." Journal of Soft Computing Paradigm 7, no. 2 (2025): 124-143.