Brain Tumor Image Segmentation with Convolutional Neural Networks: A Review

Beibei Hou; Tiansong Sheng

doi:10.54097/1na3rw11

Authors

Beibei Hou
Tiansong Sheng

DOI:

https://doi.org/10.54097/1na3rw11

Keywords:

Brain tumor segmentation, Deep Learning, Medical image analysis, BraTS Dataset, LKC

Abstract

Brain tumor segmentation is essential in medical image analysis for clinical diagnosis, treatment planning, and prognosis. Despite significant progress, challenges remain, including limited data annotation, high computational costs, and poor model generalization. To address these, researchers have proposed CNN-based models (e.g., FCN, U-Net, U-Net++) and advanced architectures like large kernel convolution (LKC), deformable convolution (DCN), and CNN-transformer hybrids. This paper examines the widely used BraTS dataset and evaluation metrics such as Dice coefficients and Hausdorff distances, while addressing current challenges. Researchers are also exploring strategies like joint learning, self-supervised learning, multimodal fusion, and lightweight model design. These advances aim to improve segmentation performance and expand clinical applications.

Downloads

Download data is not yet available.

References

[1] Louis DN, Perry A, Reifenberger G, et al. The 2016 World Health Organization Classification of Tumors of the Central Nervous System: A Summary. Acta Neuropathol. 2016, Vol. 131 (No. 6), p. 803–820.

[2] Aboussaleh I, Riffi J, El Fazazy K, et al. 3DUV-NetR+: A 3D hybrid semantic architecture using transformers for brain tumor segmentation with multimodal MR images. Results Eng. 2024, Vol. 21, p. 101892.

[3] Bahadure NB, Ray AK, Thethi HP, et al. Comparative approach of MRI-based brain tumor segmentation and classification using genetic algorithm. J Digit Imaging. 2018, Vol. 31 (No. 4), p. 477–489.

[4] Benabid A, Yuan J, Elhassan MAM, et al. CFNet: Cross-scale fusion network for medical image segmentation. J King Saud Univ Comput Inf Sci. 2024, Vol. 36 (No. 7), p. 102123.

[5] Al-Zoghby AM, Al-Awadly EMK, Moawad A, et al. Dual Deep CNN for Tumor Brain Classification. Diagnostics. 2023, Vol. 13 (No. 12), p. 2050.

[6] Kamnitsas K, Ledig C, Newcombe VFJ, et al. Efficient multi-scale 3D CNN with fully connected CRF for accurate brain lesion segmentation. Med Image Anal. 2017, Vol. 36, p. 61–78.

[7] Long J, Shelhamer E, Darrell T, et al. Fully convolutional networks for semantic segmentation. IEEE Conf Comput Vis Pattern Recognit. Boston, USA, 2015, p. 3431–3440.

[8] Ronneberger O, Fischer P, Brox T, et al. U-Net: Convolutional Networks for Biomedical Image Segmentation. Med Image Comput Comput Assist Interv. Munich, Germany, 2015, p. 234–241.

[9] Liu Z, Tong L, Chen L, et al. Deep learning based brain tumor segmentation: a survey. Complex Intell Syst. 2022, Vol. 8 (No. 4), p. 1–27.

[10] Wang H, Cao P, Wang J, et al. UCTransNet: Rethinking the Skip Connections in U-Net from a Channel-wise Perspective with Transformer [J]. arXiv preprint arXiv:2109.04335, 2021.

[11] Zhou Z, Siddiquee MMR, Tajbakhsh N, et al. UNet++: A nested U-Net architecture for medical image segmentation. Lecture Notes in Computer Science. Springer, 2018, Vol. 11045, p. 3–11.

[12] Liu Z, Mao H, Wu C-Y, et al. ConvNeXt: Revisiting Convolutions for Image Recognition. IEEE Conf Comput Vis Pattern Recognit. New Orleans, USA, 2022, p. 4929–4939.

[13] Peng C, Zhang X, Yu G, et al. Large kernel matters—improve semantic segmentation by global convolutional network [J]. arXiv preprint arXiv:1703.02719, 2017.

[14] Azad R, Niggemeier L, Hüttemann M, et al. Beyond self-attention: deformable large kernel attention for medical image segmentation [J]. arXiv preprint arXiv:2309.00121, 2023.

[15] Liu B, Wang M, Foroosh H, et al. Sparse convolutional neural networks. IEEE Conf Comput Vis Pattern Recognit. Boston, USA, 2015, p. 806–814.