A Comparative Study of 2D and 3D U-Net-Based Architectures for Pancreas and Tumor Segmentation in CT Imaging

Öz

Pancreatic cancer ranks among the deadliest forms of cancer, largely due to its asymptomatic progression, late-stage diagnosis, and the pancreas’s deep-seated and anatomically complex location within the abdominal cavity. These factors collectively prevent early detection, which is critical for improving patient outcomes. Deep learning-based segmentation models, particularly those based on the U-Net architecture, have demonstrated strong potential for the automatic identification of pancreas and pancreatic tumors in computed tomography (CT) images. This study presents a comparative analysis of 2D and 3D U-Net architectures for the segmentation of the pancreas and its tumors. The results show that the 3D Res-U-Net achieved the best performance, with a Dice score of 0.81 for pancreas segmentation and 0.51 for tumor segmentation, outperforming both the 2D U-Net and the 3D U-Net. Additionally, inference time analysis for real-time clinical settings showed that the 2D U-Net enables the fastest predictions, whereas the 3D U-Net and 3D Res-U-Net require significantly higher computation, underscoring key trade-offs for clinical deployment. These findings offer valuable insights into model selection and optimization strategies for real-world clinical applications. The findings underscore the feasibility of developing efficient deep learning-based systems to support early diagnosis and personalized treatment planning for pancreatic cancer.

Anahtar Kelimeler

• 3D U-Net
• 3D-Res-U-Net
• CT imaging
• deep learning
• medical image segmentation
• pancreatic cancer
• U-Net

Destekleyen Kurum

AKGUN Computer Incorporated Company

Etik Beyan

There are no ethical issues after the publication of this manuscript.

Teşekkür

We gratefully acknowledge the AKGÜN Technology for their contributions and financial support.

Kaynakça

1. [1].Cancer Research UK. Pancreatic cancer survival. https://www.cancerresearchuk.org/about-cancer/pancreatic-cancer/survival (accessed at 30.04.2025).
2. [2]. Isensee, F., Jaeger, P. F., Kohl, S. A., Petersen, J., & Maier-Hein, K. H. (2021). nnU-Net: A self-configuring method for deep learning-based biomedical image segmentation. Nature Methods, 18(2), 203–211. (https://doi.org/10.1038/s41592-020-01008-z )
3. [3]. Christ, P. F., Ettlinger, F., Grün, F., Elshaera, M. E. A., Lipkova, J., Schlecht, S., & Menze, B. (2017). Automatic liver and tumor segmentation of CT and MRI volumes using cascaded fully convolutional neural networks. arXiv preprint, arXiv:1702.05970.
4. [4]. Bozkır, Ö. F., Urfalı, A., Çelikten, A., Demirel, S., Budak, A., Karataş, H., & Ceylan, M. (2024). Enhancing aorta segmentation in contrast CT images: A novel deep architectural approach. Osmaniye Korkut Ata University Journal of The Institute of Science and Technology, 7(5), 2284–2303. (https://doi.org/10.47495/okufbed.1457674 )
5. [5]. Zhang, R., Zheng, Y., Poon, C. C. Y., Shen, D., & Lau, J. Y. W. (2018). Polyp detection during colonoscopy using a regression-based convolutional neural network with a tracker. Pattern Recognition, 83, 209–219.
6. (https://doi.org/10.1016/j.patcog.2018.05.026) [6]. Jain, S., Sikka, G., & Dhir, R. (2024). A systematic literature review on pancreas segmentation from traditional to non-supervised techniques in abdominal medical images. Artificial Intelligence Review, 57(12), 317. (https://doi.org/10.1007/s10462-024-10966-1)
7. [7]. Li, M., Lian, F., Li, Y., & Guo, S. (2022). Attention-guided duplex adversarial U-Net for pancreatic segmentation from computed tomography images. Journal of Applied Clinical Medical Physics, 23(4), e13537. (https://doi.org/10.1002/acm2.13537)
8. [8]. Devendhar, T., Priya, L., & Jayalakshmy, S. (2024). U-Net based pancreas segmentation from computed tomography images. In Proceedings of the 2024 Third International Conference on Smart Technologies and Systems for Next Generation Computing (ICSTSN), pp. 1–5. IEEE. (https://doi.org/10.1109/ICSTSN61422.2024.10671042)

9. [9]. Ghorpade, H., Kolhar, S., Jagtap, J., & Chakraborty, J. (2024). An optimized two stage U-Net approach for segmentation of pancreas and pancreatic tumor. MethodsX, 13, 102995. (https://doi.org/10.1016/j.mex.2024.102995)
10. [10]. Tualombo, M. E., Reyes, I., Vizcaino-Imacaña, P., & Morocho-Cayamcela, M. E. (2024). Pancreas segmentation using SRGAN combined with U-Net neural network. In Conference on Information and Communication Technologies of Ecuador, pp. 21–38. Springer Nature Switzerland. (https://doi.org/10.1007/978-3-031-75431-9_2)
11. [11]. Bayram, A. F., Gürkan, C., Budak, A., & Karataş, H. (2022). Fully automatic end-to-end convolutional neural networks-based pancreatic tumor segmentation on CT modality. Turkish Journal of Forecasting, 6(2), 67–72. (https://doi.org/10.34110/forecasting.1190299)
12. [12]. Derin, A., Gürkan, C., Budak, A., & Karataş, H. (2022). Pancreas segmentation using U-Net based segmentation networks in CT modality: A comparative analysis. European Journal of Science and Technology, 40, 94–98. (https://doi.org/10.31590/ejosat.1171803)
13. [13]. Chen, L., & Wan, L. (2023). CTUNet: Automatic pancreas segmentation using a channel-wise transformer and 3D U-Net. The Visual Computer, 39(11), 5229–5243. (https://doi.org/10.1007/s00371-022-02656-2)
14. [14]. Uslucuk, A., & Öcal, H. (2023). Comparative analysis of baseline VNet and UNet architectures on pancreas segmentation. Orclever Proceedings of Research and Development, 3(1), 146–157. (https://doi.org/10.56038/oprd.v3i1.309)
15. [15]. Ghorpade, H., Jagtap, J., Patil, S., Kotecha, K., Abraham, A., Horvat, N., & Chakraborty, J. (2023). Automatic segmentation of pancreas and pancreatic tumor: A review of a decade of research. IEEE Access, 11, 108727–108745. (https://doi.org/10.1109/ACCESS.2023.3320570)
16. [16]. Antonelli, M., Reinke, A., Bakas, S., Farahani, K., Kopp-Schneider, A., Landman, B. A., & Cardoso, M. J. (2022). The medical segmentation decathlon. Nature Communications, 13(1), 4128. (https://doi.org/10.1038/s41467-022-30695-9)
17. [17]. Ji, Y., Bai, H., Yang, J., Ge, C., Zhu, Y., Zhang, R., Li, Z., Zhang, L., Ma, W., Wan, X., & Luo, P. (2022). AMOS: A large-scale abdominal multi-organ benchmark for versatile medical image segmentation. Advances in Neural Information Processing Systems, 35, 36722–36732.(https://arxiv.org/abs/2206.08023)
18. [18]. Wasserthal, J., Breit, H. C., Meyer, M. T., Pradella, M., Hinck, D., Sauter, A. W., Heye, T., Boll, D., Cyriac, J., Yang, S., Bach, M., & Segeroth, M. (2023). TotalSegmentator: Robust segmentation of 104 anatomic structures in CT images. Radiology: Artificial Intelligence, 5(5), e230024.(https://doi.org/10.1148/ryai.230024)
19. [19]. Zhang, U. (2021). Pancreas cancer segmentation on portal-venous phase CTs [Dataset]. Zenodo. (https://doi.org/10.5281/zenodo.5510650)
20. [20]. Ma, J., et al. (2021). FLARE 2021: Fast and low-resource abdominal organ segmentation challenge [Dataset]. Grand Challenge. https://flare.grand-challenge.org/ (accessed at 23.06.2025).
21. [21]. Ronneberger, O., Fischer, P., & Brox, T. (2015). U-Net: Convolutional networks for biomedical image segmentation. In Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015, pp. 234–241. Springer International Publishing. (https://doi.org/10.1007/978-3-319-24574-4_28)
22. [22]. Çiçek, Ö., Abdulkadir, A., Lienkamp, S. S., Brox, T., & Ronneberger, O. (2016). 3D U-Net: Learning dense volumetric segmentation from sparse annotation. In Medical Image Computing and Computer-Assisted Intervention – MICCAI 2016, pp. 424–432. Springer International Publishing. (https://doi.org/10.1007/978-3-319-46723-8_49)
23. [23]. Fu, W., Ma, Z., Yang, Z., Yu, S., Zhang, Y., Zhang, X., Mei, B., Meng, Y., Ma, C., & Gong, X. (2025). Fully automated image quality assessment based on deep learning for carotid computed tomography angiography: A multicenter study. Physica Medica, 134, 104990. (https://doi.org/10.1016/j.ejmp.2025.104990)
24. [24]. Wang, J., Zhang, J. H., Zhang, J. L., Lu, F. M., Meng, R. G., & Wang, Z. (2021). Research on fault recognition method combining 3D Res-UNet and knowledge distillation. Applied Geophysics, 18(2), 199–212. (https://doi.org/10.1007/s11770-021-0894-2)
25. [25]. Liu, J., Yildirim, O., Akin, O., & Tian, Y. (2023). AI-driven robust kidney and renal mass segmentation and classification on 3D CT images. Bioengineering, 10(1), 116. (https://doi.org/10.3390/bioengineering10010116)
26. [26]. Cao, L., & Li, J. (2024). Strongly representative semantic-guided segmentation network for pancreatic and pancreatic tumors. Biomedical Signal Processing and Control, 87, 105562. (https://doi.org/10.1016/j.bspc.2023.105562)