Segmentation of liver volumetric lesions in multiphase CT images using the nnU-Net framework

The article presents a study on the application of the nnU-Net (v2) framework for automatic segmentation and classification of liver space-occupying lesions on abdominal computed tomography. The main attention is paid to the effect of the batch size and the use of data from different contrast phases on the classification accuracy of such lesions as cysts, hemangiomas, carcinomas, and focal nodular hyperplasia (FNH). During the experiments, batch sizes of 2, 3, and 4 were used, as well as data from two contrast phases ‒ arterial and venous. The results showed that the optimal batch size is 3 or 4, depending on the pathology, and the use of data from two contrast phases significantly improves the accuracy and sensitivity of space-occupying lesions classification, especially for carcinomas and cysts. The achieved best sensitivity rates were 100% for carcinomas, 94% for cysts, 81% for hemangiomas, and 84% for FNH. The paper confirms the effectiveness of nnU-Net v2 for solving medical image segmentation and classification problems and highlights the importance of choosing the right training parameters and data to achieve the best results in medical diagnostics.

1. Sato M., Jin Z., Suzuki K. Semantic Segmentation of Liver Tumor in Contrast-enhanced Hepatic CT by Using Deep Learning with Hessian-based Enhancer with Small Training Dataset Size. In: 2021 IEEE 18th International Symposium on Biomedical Imaging, 13–16 April 2021, Nice, France. IEEE; 2021. pp. 34–37. https://doi.org/10.1109/ISBI48211.2021.9433929

2. Dostovalova A.M., Gorshenin A.K., Starichkova Ju.V., Arzamasov K.M. Comparative analysis of modifications of U-Net neuronal network architectures in medical image segmentation. Digital Diagnostics. 2024;5(4):833–853. (In Russ.). https://doi.org/10.17816/DD629866

3. Huang W., Liu W., Zhang X., et al. LIDIA: Precise Liver Tumor Diagnosis on Multi-Phase Contrast-Enhanced CT via Iterative Fusion and Asymmetric Contrastive Learning. In: Medical Image Computing and Computer Assisted Intervention – MICCAI: 27th International Conference: Proceedings, Part IX, 06–10 October 2024, Marrakesh, Morocco. Cham: Springer; 2024. pp. 394–404. https://doi.org/10.1007/978-3-031-72114-4_38

4. Isensee F., Jaeger P.F., Kohl S.A.A., Petersen J., Maier-Hein K.H. nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation. Nature Methods. 2021;18:203–211. https://doi.org/10.1038/s41592-020-01008-z

5. Isensee F., Wald T., Ulrich C., et al. nnU-Net Revisited: A Call for Rigorous Validation in 3D Medical Image Segmentation. In: Medical Image Computing and Computer Assisted Intervention – MICCAI: 27th International Conference: Proceedings, Part IX, 06–10 October 2024, Marrakesh, Morocco. Cham: Springer; 2024. pp. 488–498. https://doi.org/10.1007/978-3-031-72114-4_47

6. Ronneberger O., Fischer Ph., Brox Th. U-Net: Convolutional Networks for Biomedical Image Segmentation. In: Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015: 18th International Conference: Proceedings, Part III, 05–09 October, 2015, Munich, Germany. Cham: Springer; 2015. pp. 234–241. https://doi.org/10.1007/978-3-319-24574-4_28

7. Wang J., Zhang X., Guo L., Shi Ch., Tamura Sh. Multi-scale attention and deep supervision-based 3D UNet for automatic liver segmentation from CT. Mathematical Biosciences and Engineering. 2023;20(1):1297–1316. https://doi.org/10.3934/mbe.2023059

8. He K., Zhang X., Ren Sh., Sun J. Deep Residual Learning for Image Recognition. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 27–30 June 2016, Las Vegas, NV, USA. IEEE; 2016. pp. 770–778. https://doi.org/10.1109/CVPR.2016.90

9. Kovun V.A., Kashirina I.L. Usage of U-Net and W-Net neural network architectures for steel samples metallographic analysis. Proceedings of Voronezh State University. Series: Systems Analysis and Information Technologies. 2022;(1):101–110. (In Russ.). https://doi.org/10.17308/sait.2022.1/9205

10. Wang W., Wang F., Chen Q., et al. Phase Attention Model for Prediction of Early Recurrence of Hepatocellular Carcinoma With Multi-Phase CT Images and Clinical Data. Frontiers in Radiology. 2022;2. https://doi.org/10.3389/fradi.2022.856460