Fetal Brain Anomaly Detection via Ultrasound Imaging Using Traditional and Separable CNNs with Xception
Keywords:
N\AAbstract
Delayed or missed diagnosis of fetal brain anomalies during pregnancy can result in significant developmental challenges and increased rates of newborn death. To address this critical issue, this investigation introduces a diagnostic aid powered by deep learning for the initial detection of these abnormalities through ultrasound imaging, a method recognized for its broad availability and affordability. In this study, we undertook a comparative evaluation of three different convolutional neural network (CNN) architectures, a fundamental CNN framework, a CNN integrated with depthwise separable convolutions, and the Xception model. Our investigation utilized a dataset of 1,786 meticulously labeled ultrasound images, which encompassed 16 varied categories of fetal brain anomalies, including instances of arnold-chiari malformation, ventriculomegaly ranging in severity, and intracranial tumours. After training each model for five epochs, the Xception network demonstrated the highest degree of precision in anomaly classification. Additionally, a graphical user interface (GUI) was developed to enable healthcare professionals to submit ultrasound images and obtain diagnostic predictions accompanied by their respective confidence scores. The outcomes suggest that even with limited training iterations, deep learning models exhibit a notable capability to discern intricate features within fetal ultrasound imagery.
Downloads
Metrics
References
M. M. Saleem, G. A. A. Ibrahim, F. Alomran, and M. A. Zain, “Fetal brain anomalies: sonographic features and perinatal outcomes,” BMC Pediatrics, vol. 18, no. 1, pp. 1–7, 2018. [Online]. Available: https://bmcpediatr.biomedcentral.com/articles/10.1186/s12887-018-1149-0
A. Kumar and D. Dutta, “Detection of brain anomalies in fetus using deep learning,” in Intelligent Data Communication Technologies and Internet of Things (ICICI 2021), vol. 194, X. Li, A. Balasubramanian, and V. E. Balas, Eds. Singapore: Springer, 2021, pp. 61–73. [Online]. Available: https://link.springer.com/chapter/10.1007/978-981-16-1554-2_6
World Health Organization, Maternal and perinatal health statistics. Geneva: WHO, 2020. [Online]. Available: https://www.who.int/data/gho/data/themes/topics/maternal-and-perinatal-health
P. D. Griffiths, C. Bradburn, and P. Campbell, “Magnetic resonance imaging and ultrasound of the fetal central nervous system,” European Journal of Radiology, vol. 82, no. 6, pp. 834–841, Jun. 2013, doi: 10.1016/j.ejrad.2012.12.018.
W. Xie, E. Noble, and J. A. Noble, “Automatic fetal brain extraction from ultrasound images using convolutional neural networks,” Ultrasound in Obstetrics & Gynecology, vol. 58, no. 4, pp. 600–607, 2021, doi: 10.1002/uog.21967.
S. Mhatre and J. Bakal, “Classification of fetal brain abnormalities in ultrasound images using deep convolutional neural networks,” in Proc. 2020 6th Int. Conf. on Computing Communication and Automation (ICCCA), Greater Noida, India, 2020, pp. 1–5, doi: 10.1109/ICCCA49541.2020.9250869.
F. Chollet, “Xception: Deep learning with depthwise separable convolutions,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), 2017, pp. 1251–1258, doi: 10.1109/CVPR.2017.195.
Y. Cai and M. Zhao, “A cascade convolutional neural network for fetal brain MRI segmentation,” IEEE Access, vol. 9, pp. 134497–134508, 2021, doi: 10.1109/ACCESS.2021.3117009.
H. Trivedi, “Fetal Brain Abnormalities Ultrasound Dataset,” Roboflow Universe. [Online]. Available: https://universe.roboflow.com/hritwik-trivedi-gkgrv/fetal-brain-abnormalities-ultrasound. [Accessed: May 2, 2025].
N. Avisdris et al., “Automatic linear measurements of the fetal brain on MRI with deep neural networks,” Int. J. Comput. Assist. Radiol. Surg., vol. 16, no. 9, pp. 1481–1492, 2021. doi: 10.1007/s11548-021-02436-8
R. Gu et al., “CA-Net: Comprehensive attention convolutional neural networks for explainable medical image segmentation,” IEEE Trans. Med. Imaging, vol. 40, no. 2, pp. 699–711, 2021. doi: 10.1109/TMI.2020.
R. Faghihpirayesh, D. Karimi, D. Erdoğmuş, and A. Gholipour, “Deep learning framework for real-time fetal brain segmentation in MRI,” in Proc. Int. Workshop Preterm, Perinatal and Paediatric Image Analysis, 2022, pp. 60–70. doi: 10.1007/978-3-031-17991-0_6.
[26] M. B. M. Ranzini, L. Fidon, S. Ourselin, M. Modat, and T. Vercauteren, “MONAIfbs: MONAI-based fetal brain MRI deep learning segmentation,” arXiv preprint, arXiv:2103.13314, 2021. [Online]. Available: https://arxiv.org/abs/2103.13314arXiv
C. B. N. Tran et al., “Development of gestational age–based fetal brain and intracranial volume reference norms using deep learning,” American Journal of Neuroradiology, vol. 43, no. 1, pp. 1–8, 2022. doi: 10.3174/ajnr.A7747.
R. Sreelakshmy, R. S. Rajesh, and M. S. Rajasree, “An automated deep learning model for the cerebellum segmentation from fetal brain images,” Computers in Biology and Medicine, vol. 144, p. 105391, 2022, doi: 10.1016/j.compbiomed.2022.105391.PubMed+1PMC+1
D. Karimi et al., “Detailed delineation of the fetal brain in diffusion MRI via multi-task learning,” arXiv preprint, arXiv:2409.06716, 2024. [Online]. Available: https://arxiv.org/abs/2409.06716arXiv
Downloads
Published
How to Cite
Issue
Section
License
This work is licensed under a Creative Commons Attribution 4.0 International License.
You are free to:
- Share — copy and redistribute the material in any medium or format
- Adapt — remix, transform, and build upon the material for any purpose, even commercially.
Terms:
- Attribution — You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use.
- No additional restrictions — You may not apply legal terms or technological measures that legally restrict others from doing anything the license permits.
