Improved Brain Hemorrhage Segmentation Using an Optimized U-Net With Residual Blocks
VERSION OF RECORD ONLINE: 09/09/2025
Corressponding author's email:
dothanhhieukt@gmail.comDOI:
https://doi.org/10.54644/jte.2025.1968Keywords:
Intracranial hemorrhage segmentation, Computed Tomography, Convolutional Neural Network, U-Net, ResUNet, Graphics processing unitAbstract
Brain hemorrhage is a life‑threatening emergency that demands rapid and accurate diagnosis for timely treatment. Currently, computed tomography (CT) is the primary imaging modality, but traditional segmentation techniques still exhibit limited accuracy. In this study, three variants of the U‑Net architecture—U‑Net integrated with VGG16, U‑Net integrated with ResNet‑18, and Residual U‑Net—were compared to identify the optimal solution. Notably, the Residual U‑Net leverages shortcut connections and residual blocks to improve deep learning performance even when training data are scarce. To balance background and hemorrhage segmentation, the training process employs a combined loss function of Binary Cross Entropy and Dice Loss. Evaluation results demonstrate that the Residual U‑Net outperforms the other two variants in accuracy and key performance metrics, even on small datasets. Thanks to its effective feature reuse and optimized loss function, the Residual U‑Net shows great promise as a powerful clinical support tool for enhancing the effectiveness of cerebral hemorrhage segmentation in CT images.
Downloads: 0
References
P. D. Chang et al., “Hybrid 3D/2D convolutional neural network for hemorrhage evaluation on head CT,” Am. J. Neuroradiol., vol. 39, no. 9, pp. 1609–1616, 2018.
P. Inkeaw et al., “Automatic hemorrhage segmentation on head CT scan for traumatic brain injury using 3D deep learning model,” Comput. Biol. Med., vol. 146, 2022.
K. Jnawali, M. R. Arbabshirani, N. Rao, and A. A. Patel, “Deep 3D convolution neural network for CT brain hemorrhage classification,” in Proc. SPIE Med. Imaging: Comput.-Aided Diagnosis, 2018, vol. 10575, Art. no. 105751C.
H. Ye et al., “Precise diagnosis of intracranial hemorrhage and subtypes using a three-dimensional joint convolutional and recurrent neural network,” Eur. Radiol., vol. 29, pp. 6191–6201, 2022.
M. Grewal, M. M. Srivastava, P. Kumar, and S. Varadarajan, “RADnet: Radiologist level accuracy using deep learning for hemorrhage detection in CT scans,” in Proc. IEEE Int. Symp. Biomed. Imaging (ISBI), Washington, DC, USA, 2018, pp. 281–284, doi: 10.1109/ISBI.2018.8363574.
O. Ronneberger, P. Fischer, and T. Brox, “U-Net: Convolutional networks for biomedical image segmentation,” in Proc. Med. Image Comput. Comput.-Assist. Intervent. (MICCAI), 2015.
H. S. Bhadauria, A. Singh, and M. L. Dewal, “An integrated method for hemorrhage segmentation from brain CT imaging,” Comput. Electr. Eng., vol. 39, no. 5, pp. 1527–1536, 2013.
M. K. Nag, S. Chatterjee, A. K. Sadhu, J. Chatterjee, and N. Ghosh, “Computer-assisted delineation of hematoma from CT volume using autoencoder and Chan Vese model,” Int. J. Comput. Assist. Radiol. Surg., vol. 14, pp. 259–269, 2019.
H. Kuang, B. K. Menon, and W. Qiu, “Segmenting hemorrhagic and ischemic infarct simultaneously from follow-up non-contrast CT images in patients with acute ischemic stroke,” in Lecture Notes in Computer Science, vol. 11766, pp. 856–863, 2019.
M. D. Hssayeni, M. S. Croock, A. D. Salman, H. F. Al-khafaji, Z. A. Yahya, and B. Ghoraani, “Intracranial hemorrhage segmentation using a deep convolutional model,” Data, vol. 5, no. 1, p. 14, 2020.
Q. T. Hoang, X. H. Pham, X. T. Trinh, A. V. Le, M. V. Bui, and T. T. Bui, “An efficient CNN-based method for intracranial hemorrhage segmentation from computerized tomography imaging,” J. Imaging, vol. 10, no. 4, p. 77, Mar. 25, 2024.
PhysioNet, “CT Intracranial Hemorrhage dataset (version 1.3.1),” [Online]. Available: https://physionet.org/content/ct-ich/1.3.1
S. Liu and W. Deng, “Very deep convolutional neural network based image classification using small training sample size,” in Proc. IAPR Asian Conf. Pattern Recognit. (ACPR), Kuala Lumpur, Malaysia, 2015, pp. 730–734.
M. Shafiq and Z. Gu, “Deep residual learning for image recognition: A survey,” Appl. Sci., vol. 12, p. 8972, 2022.
P. Yakubovskiy, segmentation_models (Version 1.0.1) [Software]. GitHub, 2022. [Online]. Available: https://github.com/qubvel/segmentation_models
Downloads
Published
How to Cite
Issue
Section
Categories
License
Copyright (c) 2006 Journal of Technical Education Science
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
Copyright © JTE.
