Deep Learning-Based Classification of Bladder Cancer Using Vision Transformers: A Comparative Study of ViT Models
Abstract views: 33
/ 
PDF downloads: 19
Keywords:
Vision Transfomer, Deep Learning, Bladder CancerAbstract
Bladder cancer is one of the most common cancer types of the urogenital system. Each year,
approximately 350,000 new cases are diagnosed, resulting in 150,000 deaths. Early detection of bladder
cancer plays a critical role in determining treatment strategies and reducing mortality rates. Therefore, the
development of more effective diagnostic and therapeutic approaches for bladder cancer is of significant
importance. Based on its invasion of muscle tissue, bladder cancer can develop in two distinct forms:
Non-Muscle-Invasive Bladder Cancer (NMIBC) and Muscle-Invasive Bladder Cancer (MIBC). NMIBC
is an early-stage cancer type where the cancer is confined to the surface of the bladder without invading
the muscle layer. In contrast, MIBC is a more advanced and dangerous type of cancer that invades
surrounding tissues. This study proposes an autonomous system based on the deep learning Vision
Transformer (ViT) model for the early detection of bladder cancer. Using an open-access, multicenter
dataset, the study compares two models to classify magnetic resonance imaging (MRI) scans of bladder
cancer. Following preprocessing of the bladder MRI images, model training was conducted to determine
the class of the data using the ViT approach. The study evaluates the performance of two ViT models,
ViT-Small Patch32 and ViT-Large Patch32, in the task of bladder cancer classification. The results of
both models were assessed using the metrics of F1-Score, Recall, Precision, and Accuracy. The study
findings reveal that the ViT-Large Patch32 model achieved a performance of 97% across all metrics,
providing more accurate and reliable results for bladder cancer classification. The proposed study is
expected to serve as a robust tool to assist experts in classifying bladder cancer and optimizing treatment
processes through its supportive mechanism during the decision-making phase.ve tedavi süreçlerinin
optimize edilmesinde güçlü bir araç sunması beklenmektedir.
Downloads
References
Jemal A, Bray F, Center MM, et al. Global cancer statstcs. CA Cancer J Cln 61: 69-90, 2011
Jung, I., & Messing, E. (2000). Molecular mechanisms and pathways in bladder cancer development and progression. Cancer control, 7(4), 325-334.
Xu, X., Zhang, X., Tian, Q., Wang, H., Cui, L. B., Li, S., ... & Liu, Y. (2019). Quantitative identification of nonmuscle‐invasive and muscle‐invasive bladder carcinomas: a multiparametric MRI radiomics analysis. Journal of Magnetic Resonance Imaging, 49(5), 1489-1498.
Deep learning on enhanced CT images can predict the muscular invasiveness of bladder cancer. Frontiers in Oncology.
Sylvester, R. J., Van Der Meijden, A. P., Oosterlinck, W., Witjes, J. A., Bouffioux, C., Denis, L., ... & Kurth, K. (2006). Predicting recurrence and progression in individual patients with stage Ta T1 bladder cancer using EORTC risk tables: a combined analysis of 2596 patients from seven EORTC trials. European urology, 49(3), 466-477.
Sherif, A., Jonsson, M. N., & Wiklund, N. P. (2007). Treatment of muscle-invasive bladder cancer. Expert Review of Anticancer Therapy, 7(9), 1279-1283.
Lucca, I., de Martino, M., Klatte, T., & Shariat, S. F. (2015). Novel biomarkers to predict response and prognosis in localized bladder cancer. Urologic Clinics, 42(2), 225-233.
Burger, M., Catto, J. W., Dalbagni, G., Grossman, H. B., Herr, H., Karakiewicz, P., ... & Lotan, Y. (2013). Epidemiology and risk factors of urothelial bladder cancer. European urology, 63(2), 234-241.
Litjens, G., Kooi, T., Bejnordi, B. E., Setio, A. A. A., Ciompi, F., Ghafoorian, M., ... & Sánchez, C. I. (2017). A survey on deep learning in medical image analysis. Medical image analysis, 42, 60-88.
Schlemper, J., Caballero, J., Hajnal, J. V., Price, A., & Rueckert, D. (2017). A deep cascade of convolutional neural networks for MR image reconstruction. In Information Processing in Medical Imaging: 25th International Conference, IPMI 2017, Boone, NC, USA, June 25-30, 2017, Proceedings 25 (pp. 647-658). Springer International Publishing.
Zuo, S., Xiao, Y., Chang, X., & Wang, X. (2022). Vision transformers for dense prediction: A survey. Knowledge-Based Systems, 253, 109552.
Fuster, S., Khoraminia, F., Kiraz, U., Kanwal, N., Kvikstad, V., Eftestøl, T., ... & Engan, K. (2022, June). Invasive cancerous area detection in Non-Muscle invasive bladder cancer whole slide images. In 2022 IEEE 14th Image, Video, and Multidimensional Signal Processing Workshop (IVMSP) (pp. 1-5). IEEE
Khosravi, P., Kazemi, E., Imielinski, M., Elemento, O., & Hajirasouliha, I. (2018). Deep convolutional neural networks enable discrimination of heterogeneous digital pathology images. EBioMedicine, 27, 317-328.
Yin PN, Kc K, Wei S, Yu Q, Li R, Haake A.R, ... Cui F. Histopathological distinction of noninvasive and invasive bladder cancers using machine learning approaches. BMC medical informatics and decision making. 2020; 20: 111
Jansen, I., Lucas, M., Bosschieter, J., de Boer, O. J., Meijer, S. L., van Leeuwen, T. G., ... & Savci-Heijink, C. D. (2020). Automated detection and grading of non–muscle-invasive urothelial cell carcinoma of the bladder. The American journal of pathology, 190(7), 1483-1490.
Cao, K., Zou, Y., Zhang, C., Zhang, W., Zhang, J., Wang, G., ... & Huang, B. (2024). A multicenter bladder cancer MRI dataset and baseline evaluation of federated learning in clinical application. Scientific Data, 11(1), 1147.
[https://www.kaggle.com/datasets/shirtgm/bladder-cancer-classification
Vujović, Ž. (2021). Classification model evaluation metrics. International Journal of Advanced Computer Science and Applications, 12(6), 599-606.
