Enhanced 3D DenseNet with CDC for Multimodal Brain Tumor Segmentation
Source abstract
Abstract
Precise tumor segmentation in multimodal MRI is crucial for glioma diagnosis and treatment planning; yet, deep learning models still struggle with irregular boundaries and severe class imbalance under computational constraints. An Enhanced 3D DenseNet with CDC architecture was proposed, integrating Central Difference Convolution, attention gates, and Atrous Spatial Pyramid Pooling for brain tumor segmentation on the BraTS 2023-GLI dataset. CDC layers enhance boundary sensitivity by combining intensity-level semantics and gradient-level features. Attention gates selectively emphasize relevant encoder features during skip connections, whereas the ASPP captures the multi-scale context with dilation rates. A hybrid loss function spanning three levels was introduced, consisting of a region-based Dice loss for volumetric overlap, a GPU-native 3D Sobel boundary loss for edge precision, and a class-weighted focal loss for handling class imbalance. The proposed model achieved a mean Dice score of 91.30% (ET: 87.84%, TC: 92.73%, WT: 93.34%) on the test set. Notably, these results were achieved with approximately 3.7 million parameters, representing a 17–76x reduction compared to the 50–200 million parameters required by transformer-based approaches. Enhanced 3D DenseNet with CDC architecture demonstrates that the integration of gradient-sensitive convolutions, attention mechanisms, multi-scale feature extraction, and multi-level loss optimization achieves competitive segmentation performance with significantly reduced computational requirements.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Advanced Neural Network Applications · Brain Tumor Detection and Classification · Medical Image Segmentation Techniques
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Department of Engineering, Artvin Çoruh University, 08100 Artvin, Türkiye
Graduate School of Natural and Applied Sciences, Department of Electrical-Electronics Engineering, Karadeniz Technical University, 61080 Trabzon, Türkiye
Department of Electrical-Electronics Engineering, Karadeniz Technical University, 61080 Trabzon, Türkiye
References (56)
Show source-provided references
1. Hallmarks of cancer: The next generation10.1016/j.cell.2011.02.013 ↗
2. Review of MRI-based brain tumor image segmentation using deep learning methods10.1016/j.procs.2016.09.407 ↗
3. A survey on brain tumor image analysis10.1007/s11517-023-02873-4 ↗
4. Whole-cell segmentation of tissue images with human-level performance using large-scale data annotations and deep learning10.1038/s41587-021-01094-0 ↗
5. Deep-learning-based brain tumor segmentation: A survey10.1007/s40747-022-00815-5 ↗
6. Ronneberger, O., Fischer, P., and Brox, T. (2015, January 5–9). U-Net: Convolutional Networks for Biomedical Image Segmentation. Proceedings of the International Conference on Medical Image Computing and Computer-Assisted Intervention, Munich, Germany.10.1007/978-3-319-24574-4_28 ↗
7. Huang, G., Liu, Z., Van Der Maaten, L., and Weinberger, K.Q. (2017, January 21–26). Densely connected convolutional networks. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, HI, USA.10.1109/CVPR.2017.243 ↗
8. Dense convolutional network and its application in medical image analysis10.1155/2022/2384830 ↗
9. Chen, J., Lu, Y., Yu, Q., Luo, X., Adeli, E., Wang, Y., Lu, L., Yuille, A.L., and Zhou, Y. (2021). TransUNetTransunet: Transformers make strong encoders for medical image segmentation. arXiv.
10. From CNNs to Transformers: A Review of Evolving Deep Learning Architectures for Brain Tumor Classification10.1109/ACCESS.2025.3625607 ↗
11. LMU-Net: Lightweight U-shaped network for medical image segmentation10.1007/s11517-023-02908-w ↗
12. A lightweight network for brain MRI segmentation10.1038/s41598-025-18062-2 ↗
13. LSMAtt-Net: A lightweight brain tumor segmentation approach with efficiency–complexity tradeoff10.1016/j.bspc.2025.108930 ↗
14. Abidin, Z.U., Naqvi, R.A., Haider, A., Kim, H.S., Jeong, D., and Lee, S.W. (2024). Recent deep learning-based brain tumor segmentation models using multi-modality magnetic resonance imaging: A prospective survey. Front. Bioeng. Biotechnol., 12.10.3389/fbioe.2024.1392807 ↗
15. Federated benchmarking of medical artificial intelligence with MedPerf10.1038/s42256-023-00652-2 ↗
16. Baid, U., Ghodasara, S., Mohan, S., Bilello, M., Calabrese, E., Colak, E., Farahani, K., Kalpathy-Cramer, J., Kitamura, F.C., and Pati, S. (2021). The rsna-asnr-miccai brats 2021 benchmark on brain tumor segmentation and radiogenomic classification. arXiv.
17. The multimodal brain tumor image segmentation benchmark (BRATS)10.1109/TMI.2014.2377694 ↗
18. Advancing the cancer genome atlas glioma MRI collections with expert segmentation labels and radiomic features10.1038/sdata.2017.117 ↗
19. Long, J., Shelhamer, E., and Darrell, T. (2015, January 7–12). Fully convolutional networks for semantic segmentation. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Boston, MA, USA.10.1109/CVPR.2015.7298965 ↗
20. Zhou, Z., Rahman Siddiquee, M.M., Tajbakhsh, N., and Liang, J. (2018, January 20). Unet++: A Nested U-Net Architecture for Medical Image Segmentation. Proceedings of the International Workshop on Deep Learning in Medical Image Analysis, Granada, Spain.10.1007/978-3-030-00889-5_1 ↗
21. Çiçek, Ö., Abdulkadir, A., Lienkamp, S.S., Brox, T., and Ronneberger, O. (2016, January 17–21). 3D U-Net: Learning Dense Volumetric Segmentation from Sparse Annotations. Proceedings of the International Conference on Medical Image Computing and Computer-Assisted Intervention, Athens, Greece.10.1007/978-3-319-46723-8_49 ↗
22. nnU-Net: A self-configuring method for deep learning-based biomedical image segmentation10.1038/s41592-020-01008-z ↗
23. Jégou, S., Drozdzal, M., Vazquez, D., Romero, A., and Bengio, Y. (2017, January 21–26). The one hundred layers tiramisu: Fully convolutional densenets for semantic segmentation. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops, Honolulu, HI, USA.10.1109/CVPRW.2017.156 ↗
24. H-DenseUNet: Hybrid densely connected UNet for liver and tumor segmentation from CT volumes10.1109/TMI.2018.2845918 ↗
25. Wu, X., Yang, X., Li, Z., Liu, L., and Xia, Y. (2024). Multimodal brain tumor image segmentation based on DenseNet. PLoS ONE, 19.10.1371/journal.pone.0286125 ↗
26. Peiris, H., Hayat, M., Chen, Z., Egan, G., and Harandi, M. (2022, January 18–22). A Robust Volumetric Transformer for Accurate 3D Tumor Segmentation. Proceedings of the International Conference on Medical Image Computing and Computer-Assisted Intervention, Singapore.10.1007/978-3-031-16443-9_16 ↗
27. Hatamizadeh, A., Nath, V., Tang, Y., Yang, D., Roth, H.R., and Xu, D. (2021, January 27). Swin Unetr: Swin Transformers for Semantic Segmentation Of Brain Tumors in Mri Images. Proceedings of the International MICCAI Brainlesion Workshop, Virtual Event.10.1007/978-3-031-08999-2_22 ↗
28. BTSwin-Unet: 3D U-shaped symmetrical Swin transformer-based network for brain tumor segmentation with self-supervised pre-training10.1007/s11063-022-10919-1 ↗
29. Xing, S., Lai, Z., Zhu, J., He, W., and Mao, G. (2025). Semantic Segmentation of Brain Tumors Using a Local–Global Attention Model. Appl. Sci., 15.10.3390/app15115981 ↗
30. Deeplab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected crfs10.1109/TPAMI.2017.2699184 ↗
31. Automatic segmentation using a hybrid dense network integrated with an 3D-atrous spatial pyramid pooling module for computed tomography (CT) imaging10.1109/ACCESS.2020.3024277 ↗
32. Echine, K., and Darouichi, A. (2024, January 22–24). 3D AGSE-Res-UNet: An Efficient Deep Learning Approach based on Res-UNet for Brain Tumor Segmentation. Proceedings of the 2024 International Conference on Connected Innovation and Technology (ICCITX), Rabat, Morocco.10.1109/ICCITX61791.2024.11070871 ↗
33. Brain tumor segmentation with advanced nnU-Net: Pediatrics and adults tumors10.1016/j.neuri.2024.100156 ↗
34. Chowdhury, S.R., and Mirzaei, G. (2025, January 14–17). Hybridization of Attention UNet with Repeated Atrous Spatial Pyramid Pooling for Improved Brain Tumor Segmentation. Proceedings of the 2025 IEEE 22nd International Symposium on Biomedical Imaging (ISBI), London, UK.10.1109/ISBI60581.2025.10980895 ↗
35. Yu, Z., Zhao, C., Wang, Z., Qin, Y., Su, Z., Li, X., Zhou, F., and Zhao, G. (2020, January 13–19). Searching for central difference convolutional networks for face anti-spoofing. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Seattle, WA, USA.10.1109/CVPR42600.2020.00534 ↗
36. UCDCN: A nested architecture based on central difference convolution for face anti-spoofing10.1007/s40747-024-01397-0 ↗
37. Ashraf, M., Alharthi, R., Chen, L., Umer, M., Alsubai, S., and Eshmawi, A.A. (2024). Attention 3D central difference convolutional dense network for hyperspectral image classification. PLoS ONE, 19.10.1371/journal.pone.0300013 ↗
38. Oktay, O., Schlemper, J., Folgoc, L.L., Lee, M., Heinrich, M., Misawa, K., Mori, K., McDonagh, S., Hammerla, N.Y., and Kainz, B. (2018). Attention u-net: Learning where to look for the pancreas. arXiv.
39. Guo, C. (2023). 3d sa-unet: 3d spatial attention unet with 3d aspp for white matter hyperintensities segmentation. arXiv.
40. Azad, R., Heidary, M., Yilmaz, K., Hüttemann, M., Karimijafarbigloo, S., Wu, Y., Schmeink, A., and Merhof, D. (2023). Loss Functions in the Era of Semantic Segmentation: A Survey and Outlook. arXiv.
41. Gammoudi, I., Ghozi, R., and Mahjoub, M.A. (2022, January 27–29). HDFU-Net: An Improved Version of U-Net using a Hybrid Dice Focal Loss Function for Multimodal Brain Tumor Image Segmentation. Proceedings of the 2022 International Conference on Cyberworlds (CW), Kanazawa, Japan.10.1109/CW55638.2022.00019 ↗
42. Combo Loss: Handling Input and Output Imbalance in Multi-Organ Segmentation10.1016/j.compmedimag.2019.04.005 ↗
43. Milletari, F., Navab, N., and Ahmadi, S.A. (2016, January 25–28). V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation. Proceedings of the 2016 Fourth International Conference on 3D Vision (3DV), Stanford, CA, USA.10.1109/3DV.2016.79 ↗
44. Boundary Loss for Highly Unbalanced Segmentation10.1016/j.media.2020.101851 ↗
45. Lin, T.Y., Goyal, P., Girshick, R., He, K., and Dollár, P. (2017, January 22–29). Focal Loss for Dense Object Detection. Proceedings of the IEEE International Conference on Computer Vision (ICCV), Venice, Italy.10.1109/ICCV.2017.324 ↗
46. N4ITK: Improved N3 bias correction10.1109/TMI.2010.2046908 ↗
47. Maani, F., Hashmi, A.U.R., Aljuboory, M., Saeed, N., Sobirov, I., and Yaqub, M. (2023). Advanced Tumor Segmentation in Medical Imaging: An Ensemble Approach for Brats 2023 Adult Glioma and Pediatric Tumor Tasks. International Challenge on Cross-Modality Domain Adaptation for Medical Image Segmentation, Springer Nature.10.1007/978-3-031-76163-8_24 ↗
48. TorchIO: A Python library for efficient loading, preprocessing, augmentation and patch-based sampling of medical images in deep learning10.1016/j.cmpb.2021.106236 ↗
49. Roy, S., Koehler, G., Ulrich, C., Baumgartner, M., Petersen, J., Isensee, F., and Maier-Hein, K.H. (2023, January 8–12). Mednext: Transformer-Driven Scaling of Convnets for Medical Image Segmentation. Proceedings of the International Conference on Medical Image Computing and Computer-Assisted Intervention, Vancouver, BC, Canada.10.1007/978-3-031-43901-8_39 ↗
50. Hatamizadeh, A., Tang, Y., Nath, V., Yang, D., Myronenko, A., Landman, B., Roth, H.R., and Xu, D. (2022, January 3–8). Unetr: Transformers for 3d medical image segmentation. Proceedings of the IEEE/CVF winter conference on applications of computer vision, Waikoloa, HI, USA.10.1109/WACV51458.2022.00181 ↗
51. UNETR++: Delving into efficient and accurate 3D medical image segmentation10.1109/TMI.2024.3398728 ↗
52. Lee, H.H., Bao, S., Huo, Y., and Landman, B.A. (2022). 3d ux-net: A large kernel volumetric convnet modernizing hierarchical transformer for medical image segmentation. arXiv.
53. Jia, H., Bai, C., Cai, W., Huang, H., and Xia, Y. (October, January 27). HNF-Netv2 for Brain Tumor Segmentation Using Multi-Modal MR Imaging. Proceedings of the International MICCAI Brainlesion Workshop, Strasbourg, France.
54. Liu, Z., Mao, H., Wu, C.Y., Feichtenhofer, C., Darrell, T., and Xie, S. (2022, January 18–24). A convnet for the 2020s. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, New Orleans, LA, USA.10.1109/CVPR52688.2022.01167 ↗
55. LATUP-Net: A lightweight 3D attention U-Net with parallel convolutions for brain tumor segmentation10.1016/j.compbiomed.2024.109353 ↗
56. Wang, T., Wu, T., Yang, D., Xu, Y., Lv, D., Jiang, T., Wang, H., Chen, Q., Xu, S., and Yan, Y. (2025). 3D-MRI brain glioma intelligent segmentation based on improved 3D U-net network. PLoS ONE, 20.10.1371/journal.pone.0325534 ↗