VGG-ViT: A Hybrid Deep Feature Framework for Zone-Specific Prostate Cancer Classification
Keywords:
Prostate cancer, Multiparametric MRI, Deep feature, Anatomical zones, ClassificationAbstract
Prostate cancer (PCa) remains a significant health concern as one of the leading causes of cancer-related mortality in men globally. The accurate differentiation between clinically significant (CS) and clinically insignificant (CiS) PCa plays a crucial role in treatment planning and patient outcomes. This research introduces an innovative framework that leverages multiparametric magnetic resonance imaging (mpMRI) for PCa classification in both the entire prostate and specific anatomical zones. Our approach combines transfer learning techniques using pre-trained VGG19 and Vision Transformer (ViT) models for feature extraction, coupled with Support Vector Machine (SVM) classification. Through exhaustive analysis of possible combinations from nine mpMRI sequences, the framework identifies optimal MRI sequences for both whole-prostate and zone-specific analysis. A thorough investigation of sequence importance is conducted by analyzing their frequency in top-performing combinations, revealing distinct patterns across different anatomical regions. The framework achieves superior classification performance with an Area Under the Curve (AUC) of 0.87 for whole-prostate assessment and notably higher performance in zone-specific assessment, with the transition zone (TZ) yielding the best performance among all analyzed zones at an AUC of 0.96. The analysis reveals the particular significance of high b-value diffusion-weighted imaging (DWI) sequences, especially BVAL (b=1400 s/mm²), followed by Ktrans imaging, in enhancing classification accuracy. These comprehensive findings provide valuable insights for optimizing zone-specific mpMRI protocols in clinical practice.
Downloads
References
F. Bray et al., “Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries,” CA: A Cancer Journal for Clinicians, vol. 74, no. 3, pp. 229–263, 2024, doi: https://doi.org/10.3322/caac.21834.
N. Mottet et al., “EAU-EANM-ESTRO-ESUR-SIOG Guidelines on Prostate Cancer—2020 Update. Part 1: Screening, Diagnosis, and Local Treatment with Curative Intent,” European Urology, vol. 79, no. 2, pp. 243–262, 2021, doi: https://doi.org/10.1016/j.eururo.2020.09.042.
B. Turkbey et al., “Prostate Imaging Reporting and Data System Version 2.1: 2019 Update of Prostate Imaging Reporting and Data System Version 2,” European Urology, vol. 76, no. 3, pp. 340–351, 2019, doi: https://doi.org/10.1016/j.eururo.2019.02.033.
J. C. Weinreb et al., “PI-RADS Prostate Imaging – Reporting and Data System: 2015, Version 2,” European Urology, vol. 69, no. 1, pp. 16–40, 2016, doi: https://doi.org/10.1016/j.eururo.2015.08.052.
H. U. Ahmed et al., “Diagnostic accuracy of multi-parametric MRI and TRUS biopsy in prostate cancer (PROMIS): a paired validating confirmatory study,” The Lancet, vol. 389, no. 10071, pp. 815–822, Feb. 2017, doi: 10.1016/S0140-6736(16)32401-1.
M. de Rooij, E. H. J. Hamoen, J. J. Fütterer, J. O. Barentsz, and M. M. Rovers, “Accuracy of Multiparametric MRI for Prostate Cancer Detection: A Meta-Analysis,” American Journal of Roentgenology, vol. 202, no. 2, pp. 343–351, Feb. 2014, doi: 10.2214/AJR.13.11046.
F.-J. H. Drost et al., “Prostate MRI, with or without MRI-targeted biopsy, and systematic biopsy for detecting prostate cancer,” Cochrane Database of Systematic Reviews, vol. 2019, no. 4, Apr. 2019, doi: 10.1002/14651858.CD012663.pub2.
A. B. Rosenkrantz et al., “Interobserver Reproducibility of the PI-RADS Version 2 Lexicon: A Multicenter Study of Six Experienced Prostate Radiologists,” Radiology, vol. 280, no. 3, pp. 793–804, 2016, doi: 10.1148/radiol.2016152542.
G. Litjens et al., “A survey on deep learning in medical image analysis,” Medical Image Analysis, vol. 42, pp. 60–88, 2017, doi: https://doi.org/10.1016/j.media.2017.07.005.
S. L. Goldenberg, G. Nir, and S. E. Salcudean, “A new era: artificial intelligence and machine learning in prostate cancer,” Nature Reviews Urology, vol. 16, no. 7, pp. 391–403, May 2019, doi: 10.1038/s41585-019-0193-3.
N. M. Al-Zidi and D. Vasumathi, “Magnetic Resonance Imaging-based Prostate Cancer Detection and Classification using Machine Learning and Deep Learning Techniques: A Survey,” in 2024 5th International Conference on Mobile Computing and Sustainable Informatics (ICMCSI), 2024, pp. 304–311. doi: 10.1109/ICMCSI61536.2024.00050.
X. Wang et al., “Searching for prostate cancer by fully automated magnetic resonance imaging classification: deep learning versus non-deep learning,” Scientific Reports, vol. 7, no. 1, p. 15415, Nov. 2017, doi: 10.1038/s41598-017-15720-y.
Q. Chen, S. Hu, P. Long, F. Lu, Y. Shi, and Y. Li, “A Transfer Learning Approach for Malignant Prostate Lesion Detection on Multiparametric MRI.,” Technology in cancer research & treatment, vol. 18, Jan. 2019, doi: 10.1177/1533033819858363.
H. J. Michaely, G. Aringhieri, D. Cioni, and E. Neri, “Current Value of Biparametric Prostate MRI with Machine-Learning or Deep-Learning in the Detection, Grading, and Characterization of Prostate Cancer: A Systematic Review,” Diagnostics, vol. 12, no. 4, p. 799, Mar. 2022, doi: 10.3390/diagnostics12040799.
Y. Zhu et al., “MRI‐based prostate cancer detection with high‐level representation and hierarchical classification,” Medical Physics, vol. 44, no. 3, pp. 1028–1039, Mar. 2017, doi: 10.1002/mp.12116.
X. Zhong et al., “Deep transfer learning-based prostate cancer classification using 3 Tesla multi-parametric MRI,” Abdominal Radiology, vol. 44, no. 6, pp. 2030–2039, Nov. 2018, doi: 10.1007/S00261-018-1824-5.
A. Dosovitskiy et al., “An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale.” 2021.
J. Chen et al., “TransUNet: Transformers Make Strong Encoders for Medical Image Segmentation,” Feb. 2021, [Online]. Available: http://arxiv.org/abs/2102.04306
C. Matsoukas, J. F. Haslum, M. Söderberg, and K. Smith, “Is it Time to Replace CNNs with Transformers for Medical Images?,” Aug. 2021, [Online]. Available: http://arxiv.org/abs/2108.09038
O. Ruprecht, P. Weisser, B. Bodelle, H. Ackermann, and T. J. Vogl, “MRI of the prostate: Interobserver agreement compared with histopathologic outcome after radical prostatectomy,” European Journal of Radiology, vol. 81, no. 3, pp. 456–460, Mar. 2012, doi: 10.1016/j.ejrad.2010.12.076.
R. Girometti et al., “Interreader agreement of PI‐RADS v. 2 in assessing prostate cancer with multiparametric MRI: A study using whole‐mount histology as the standard of reference,” Journal of Magnetic Resonance Imaging, vol. 49, no. 2, pp. 546–555, Feb. 2019, doi: 10.1002/jmri.26220.
A. S. Becker et al., “Direct comparison of PI-RADS version 2 and version 1 regarding interreader agreement and diagnostic accuracy for the detection of clinically significant prostate cancer,” European Journal of Radiology, vol. 94, pp. 58–63, Sep. 2017, doi: 10.1016/j.ejrad.2017.07.016.
R. Trigui, J. Miteran, L. Sellami, P. Walker, and A. Ben Hamida, “A classification approach to prostate cancer localization in 3T multi-parametric MRI,” in 2nd International Conference on Advanced Technologies for Signal and Image Processing, ATSIP 2016, Institute of Electrical and Electronics Engineers Inc., Jul. 2016, pp. 113–118. doi: 10.1109/ATSIP.2016.7523064.
D. Kwon et al., “Classification of suspicious lesions on prostate multiparametric MRI using machine learning.,” Journal of medical imaging (Bellingham, Wash.), vol. 5, no. 3, p. 34502, Jul. 2018, doi: 10.1117/1.JMI.5.3.034502.
A. Kitchen and J. Seah, “Support vector machines for prostate lesion classification,” in Proc.SPIE, Mar. 2017, p. 1013427. doi: 10.1117/12.2277120.
Y. Song et al., “Computer-aided diagnosis of prostate cancer using a deep convolutional neural network from multiparametric MRI,” Journal of Magnetic Resonance Imaging, vol. 48, no. 6, pp. 1570–1577, Dec. 2018, doi: 10.1002/JMRI.26047.
Y. Yuan et al., “Prostate cancer classification with multiparametric MRI transfer learning model,” Medical Physics, vol. 46, no. 2, pp. 756–765, Feb. 2019, doi: https://doi.org/10.1002/mp.13367.
M. Mehmood, S. H. Abbasi, K. Aurangzeb, M. F. Majeed, M. S. Anwar, and M. Alhussein, “A classifier model for prostate cancer diagnosis using CNNs and transfer learning with multi-parametric MRI,” Frontiers in Oncology, vol. 13, Nov. 2023, doi: 10.3389/fonc.2023.1225490.
Y. Liu and X. An, “A classification model for the prostate cancer based on deep learning,” in Proceedings - 2017 10th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics, CISP-BMEI 2017, Institute of Electrical and Electronics Engineers Inc., Feb. 2018, pp. 1–6. doi: 10.1109/CISP-BMEI.2017.8302240.
E. Pachetti and S. Colantonio, “3D-Vision-Transformer Stacking Ensemble for Assessing Prostate Cancer Aggressiveness from T2w Images,” Bioengineering, vol. 10, no. 9, 2023, doi: 10.3390/bioengineering10091015.
E. Pachetti, S. Colantonio, and M. A. Pascali, “On the Effectiveness of 3D Vision Transformers for the Prediction of Prostate Cancer Aggressiveness,” in Image Analysis and Processing. ICIAP 2022 Workshops, P. L. Mazzeo, E. Frontoni, S. Sclaroff, and C. Distante, Eds., Cham: Springer International Publishing, 2022, pp. 317–328.
A. Santhirasekaram, K. Pinto, M. Winkler, E. Aboagye, B. Glocker, and A. Rockall, “Multi-scale Hybrid Transformer Networks: Application to Prostate Disease Classification,” in Multimodal Learning for Clinical Decision Support, Cham: Springer International Publishing, 2021, pp. 12–21.
Y. Wang and M. Wang, “Selecting proper combination of mpMRI sequences for prostate cancer classification using multi-input convolutional neuronal network,” Physica Medica, vol. 80, pp. 92–100, Dec. 2020, doi: 10.1016/j.ejmp.2020.10.013.
P. Sobecki, D. Życka-Malesa, I. Mykhalevych, K. Sklinda, and A. Przelaskowski, “MRI imaging texture features in prostate lesions classification,” in IFMBE Proceedings, vol 65. Springer, 2018, pp. 827–830. doi: 10.1007/978-981-10-5122-7_207.
G. Litjens, O. Debats, J. Barentsz, N. Karssemeijer, and H. Huisman, “ProstateX Challenge data [dataset].” The Cancer Imaging Archive, 2017. doi: 10.7937/K9TCIA.2017.MURS5CL.
S. G. Armato et al., “PROSTATEx Challenges for computerized classification of prostate lesions from multiparametric magnetic resonance images,” Journal of Medical Imaging, vol. 5, no. 4, p. 44501, 2018, doi: 10.1117/1.JMI.5.4.044501.
S. Liu, H. Zheng, Y. Feng, and W. Li, “Prostate Cancer Diagnosis using Deep Learning with 3D Multiparametric MRI,” in Medical Imaging 2017: Computer-Aided Diagnosis, SPIE, Mar. 2017, p. 1013428. doi: 10.1117/12.2277121.
J. C. Y. Seah, J. S. N. Tang, and A. Kitchen, “Detection of prostate cancer on multiparametric MRI,” in Medical Imaging 2017: Computer-Aided Diagnosis, SPIE, Mar. 2017, pp. 585–588. doi: 10.1117/12.2277122.
A. Mehrtash et al., “Classification of Clinical Significance of MRI Prostate Findings Using 3D Convolutional Neural Networks.,” Proceedings of SPIE--the International Society for Optical Engineering, vol. 10134, Feb. 2017, doi: 10.1117/12.2277123.
Published
How to Cite
Issue
Section
Copyright (c) 2025 Nasser M. Al-Zidi, D. Vasumathi
This work is licensed under a Creative Commons Attribution 4.0 International License.
- Copyright and Licensing
For all articles published in SABA journals, copyright is retained by the authors. Articles are licensed under an open access Creative Commons CC BY 4.0 license, meaning that anyone may download and read the paper for free. In addition, the article may be reused and quoted provided that the original published version is cited. These conditions allow for maximum use and exposure of the work, while ensuring that the authors receive proper credit.
