IRO Journals
Journal of Soft Computing Paradigm

MRI-based Liver Tumor Image Segmentation using Deep Learning Techniques

Open Access
Published: 16 July, 2026
Pages: 253-265
  • Shobana S. , Shobana S.
    Department of Biomedical Engineering, Paavai College of Engineering, Affiliated to Anna University, Namakkal, India
    Department of Biomedical Engineering, Paavai College of Engineering, Affiliated to Anna University, Namakkal, India
  • Monisha R. , Monisha R.
    Department of Biomedical Engineering, Paavai College of Engineering, Affiliated to Anna University, Namakkal, India
    Department of Biomedical Engineering, Paavai College of Engineering, Affiliated to Anna University, Namakkal, India
  • Shanthi M. , Shanthi M.
    Department of Biomedical Engineering, Paavai College of Engineering, Affiliated to Anna University, Namakkal, India
    Department of Biomedical Engineering, Paavai College of Engineering, Affiliated to Anna University, Namakkal, India
  • Gayathri E. Gayathri E.
    Department of Biomedical Engineering, Paavai College of Engineering, Affiliated to Anna University, Namakkal, India
    Department of Biomedical Engineering, Paavai College of Engineering, Affiliated to Anna University, Namakkal, India
view PDF
view PDF

How to Cite

S., Shobana, Monisha R., Shanthi M., and Gayathri E. 2026. “MRI-Based Liver Tumor Image Segmentation Using Deep Learning Techniques”. Journal of Soft Computing Paradigm 8 (3): 253-65. https://doi.org/10.36548/jscp.2026.3.005.
ACM ACS APA ABNT Chicago Harvard IEEE MLA Turabian Vancouver AMA

Download Citation

Endnote/Zotero/Mendeley (RIS) BibTeX

Keywords

U-Net–Swin CNN
Liver Tumor Segmentation
Magnetic Resonance Imaging (MRI)
Hierarchical Feature Extraction
Tumor Localization
Quantitative Liver Analysis

Abstract

Segmentation of tumors from Magnetic Resonance Imaging (MRI) is very important for early diagnosis and treatment of the disease. In general, tumor segmentation is a challenging task due to different shapes and intensity levels of tumors, along with different liver tissues around the tumor. In this research, a hybrid network of U-net and Swin CNN has been proposed for automatic segmentation of liver tumors and quantification of tumor size using MRI images. The proposed system includes image pre-processing, hierarchical feature extraction, liver segmentation, tumor detection, and severity quantification into single pipeline. Feature extraction is done using convolutional neural network layers while long-distance contextual information is extracted using the Swin transformer blocks for enabling a better segmentation. Quantification is processed using the estimation of the liver area, tumor area, and ratio of tumor-to-liver area, and then visualization of the segmentation result. The proposed method achieves the highest accuracy of 96.5% on the MRI dataset.

References

  1. Atabansi, Chukwuemeka Clinton, Hui Li, Sheng Wang, Jing Nie, Haijun Liu, Bo Xu, Xichuan Zhou, and Dewei Li. "ICT-Net: An Integrated Convolution and Transformer-Based Network for Complex Liver and Liver Tumor Region Segmentation." IEEE Journal of Translational Engineering in Health and Medicine 2025, vol 13: 310-322.
  2. Mayuri, A. V. R., S. P. Maniraj, M. Duraisamy, G. L. N. Murthy, Kanika Garg, and M. Sangeetha. "TransDense121-UNet: a multi-scale transformer-based approach for accurate liver tumor segmentation." Evolving Systems 2025, vol 16, no. 3: 96.
  3. Song, Jian, YuChang Hu, JunHao Zhang, Rahul Jain, and Yen-Wei Chen. "Detection of focal liver lesions in CT images using a transformer-based end-to-end detection model." 13th Global Conference on Consumer Electronics (GCCE), IEEE, 2024: 1061-1064.
  4. Zhao, Ling, Shuaiqi Liu, Bing Li, Wenjia Cai, Ping Liang, Jie Yu, and Jie Zhao. "A hybrid cnn-transformer for focal liver lesion classification." International Conference on Acoustics, Speech and Signal Processing (ICASSP), IEEE, 2024: 13001-13005.
  5. Wang, Weibin, Yutaro Iwamoto, Xianhua Han, Yen-Wei Chen, Qingqing Chen, Dong Liang, Lanfen Lin, Hongjie Hu, and Qiaowei Zhang. "Classification of focal liver lesions using deep learning with fine-tuning." International Conference on Digital Medicine and Image Processing 2018, 56-60.
  6. Ma, Haiyu, and Maihemuti Maimaiti. "FUF-TransUNet: A Transformer-Based U-Net with fully utilize of features for liver and liver-tumor segmentation in CT images." In Chinese Conference on Pattern Recognition and Computer Vision (PRCV), Singapore: Springer Nature Singapore, 2024: 34-47.
  7. Kang, Ming, Chee-Ming Ting, Fung Fung Ting, and C-W. Raphaël Phan. "CAFCT-Net: A CNN-transformer hybrid network with contextual and attentional feature fusion for liver tumor segmentation." International Conference on Image Processing (ICIP), IEEE, 2024: 2970-2974.
  8. Bandaru, Sumash Chandra, G. Bharathi Mohan, R. Prasanna Kumar, and Ali Altalbe. "SwinGALE: fusion of swin transformer and attention mechanism for GAN-augmented liver tumor classification with enhanced deep learning." International Journal of Information Technology 2024, vol 16, no. 8: 5351-5369.
  9. Chen, Meiqin, Xiaoliang Jiang, Weili Lu, and Chunxian Peng. "MLFI-Net: A modified encoder-decoder network with multi-layer feature interwoven for liver tumor segmentation." IEEE Access 2025, vol 13: 161865-161880.
  10. Wu, Mian, Yinling Qian, Xiangyun Liao, Qiong Wang, and Pheng-Ann Heng. "Hepatic vessel segmentation based on 3D swin-transformer with inductive biased multi-head self-attention." BMC Medical Imaging 2023, vol 23, no. 1: 91.
  11. Viriyasaranon, Thanaporn, Sang Myung Woo, and Jang-Hwan Choi. "Unsupervised visual representation learning based on segmentation of geometric pseudo-shapes for transformer-based medical tasks." IEEE Journal of Biomedical and Health Informatics 2023, vol 27, no. 4: 2003-2014.
  12. H. Ikram, "Swin-Unet Liver Tumor Segmentation," Kaggle Notebook, 2024. Available: https://www.kaggle.com/code/hassanikram/swin-unet-liver-tumor-segmentation