Advanced Deep Learning for Stroke Classification Using Multi-Slice CT Image Analysis
Abstract
Brain stroke is a leading cause of mortality and disability globally, necessitating rapid and accurate diagnosis for timely intervention. While Computed Tomography (CT) imaging is the gold standard for stroke detection, manual interpretation is time-consuming, prone to error, and subject to inter-observer variability. Although deep learning models have shown promise in automating stroke detection, many rely on 2D analysis, ignore 3D spatial relationships, or require labour-intensive slice-level annotations, which limits their scalability and clinical applicability. To address these challenges, we propose MedHybridNet, a novel hybrid deep learning architecture that integrates convolutional neural networks (CNNs) for local feature extraction with Transformer-based modules to model global contextual dependencies across volumetric brain scans. Our main contribution is twofold: (1) the SliceAttention mechanism, which dynamically identifies diagnostically relevant slices using only patient-level labels, eliminating the need for costly slice-level annotations while enhancing interpretability through attention maps and Grad-CAM visualizations; and (2) a cGAN-based augmentation strategy that generates high-quality, pathology-informed synthetic CT slices to overcome data scarcity and class imbalance. The framework processes complete 3D brain volumes, leveraging both CNNs and Transformers in a dual-path design, and incorporates hierarchical attention for refined feature selection and classification. Evaluated via patient-wise 5-fold cross-validation on a real-world dataset of 2501 CT scans from 82 patients, MedHybridNet achieves an accuracy of 98.31%, outperforming existing methods under weak supervision. These results demonstrate its robustness, generalization capability, and superior interpretability. By combining architectural innovation with clinically relevant design choices, MedHybridNet advances the integration of Artificial Intelligence (AI) into real-world stroke care, offering a scalable, accurate, and explainable solution that can significantly improve diagnostic efficiency and patient outcomes in routine clinical practice.
Downloads
References
The top 10 causes of death. Accessed: May 13, 2024. [Online]. Available: https://www.who.int/news-room/fact-sheets/detail/the-top-10-causes-of-death
G. K. Sharma, S. Kumar, V. Ranga, and M. K. Murmu, “Artificial intelligence in cerebral stroke images classification and segmentation: A comprehensive study,” Multimedia Tools Appl., vol. 83, no. 14, pp. 43539–43575, Apr. 2024, doi: doi.org/10.1007/s11042-023-17324-3
A. Tursynova, A. Sakhipov, I. Omirzak, Z. Ikram, S. Smakova, and M. Kutubayeva, “Classification of brain strokes in computed tomography images utilizing deep learning,” in Proc. IEEE 4th Int. Conf. Smart Inf. Syst. Technol. (SIST), May 2024, pp. 328–333, doi: 10.1109/SIST61555.2024.10629580
J. Luo, P. Dai, Z. He, Z. Huang, S. Liao, and K. Liu, “Deep learning models for ischemic stroke lesion segmentation in medical images: A survey,” Comput. Biol. Med., vol. 177, p. 108509, Aug. 2024, doi: 10.1016/j.compbiomed.2024.108509
C. Huang, J. Wang, S.-H. Wang, and Y.-D. Zhang, “Applicable artificial intelligence for brain disease: A survey,” Neurocomputing, vol. 504, pp. 223–239, Sep. 2022. doi: doi.org/10.1016/j.neucom.2022.07.005
M. Arabahmadi, R. Farahbakhsh, and J. Rezazadeh, “Deep learning for smart healthcare—A survey on brain tumor detection from medical imaging,” Sensors, vol. 22, no. 5, p. 1960, Mar. 2022. doi: doi.org/10.3390/s22051960
P. Hu, T. Yan, B. Xiao, H. Shu, Y. Sheng, Y. Wu, et al., “Deep learning-assisted detection and segmentation of intracranial hemorrhage in noncontrast computed tomography scans of acute stroke patients: a systematic review and meta-analysis,” International Journal of Surgery, vol. 110, no. 6, pp. 3839–3847, 2022, doi: 10.1097/JS9.0000000000001266
S. Ahmed, J. F. Esha, M. S. Rahman, M. S. Kaiser, A. S. M. S. Hosen, D. Ghimire, and M. J. Park, “Exploring deep learning and machine learning approaches for brain hemorrhage detection,” IEEE Access, vol. 12, pp. 45060–45093, 2024. doi: 10.1109/ACCESS.2024.3376438
Z. Xu and C. Ding, “Combining convolutional attention mechanism and residual deformable transformer for infarct segmentation from CT scans of acute ischemic stroke patients,” Frontiers Neurol., vol. 14, Jul. 2023, Art. no. 1178637. doi: 10.3389/fneur.2023.1178637
M. Srikrishna et al., “Deep learning from MRI-derived labels enables automatic brain tissue classification on human brain CT,” NeuroImage, vol. 244, p. 118606, Dec. 2021, doi: doi.org/10.1016/j.neuroimage.2021.118606
U. Raghavendra, T.-H. Pham, A. Gudigar, V. Vidhya, B. N. Rao, S. Sabut, et al., “Novel and accurate non-linear index for the automated detection of haemorrhagic brain stroke using CT images,” Complex Intell. Syst., vol. 7, no. 2, pp. 929–940, Apr. 2021, doi: doi.org/10.1007/s40747-020-00257-x
A. Gautam and B. Raman, “Towards effective classification of brain hemorrhagic and ischemic stroke using CNN,” Biomed. Signal Process. Control, vol. 63, p. 102178, Jan. 2021, doi: doi.org/10.1016/j.bspc.2020.102178
R. Raj, J. Mathew, S. K. Kannath, and J. Rajan, “StrokeViT with AutoML for brain stroke classification,” Eng. Appl. Artif. Intell., vol. 119, p. 105772, Mar. 2023, doi: doi.org/10.1016/j.engappai.2022.105772
S. Yalçın and H. Vural, “Brain stroke classification and segmentation using encoder-decoder based deep convolutional neural networks,” Comput. Biol. Med., vol. 149, p. 105941, Oct. 2022, doi: 10.1016/j.compbiomed.2022.105941
O. Shafaat et al., “Leveraging artificial intelligence in ischemic stroke imaging,” J. Neuroradiol., vol. 49, no. 4, pp. 343–351, Jul. 2022, doi: 10.1016/j.neurad.2021.05.001
T. Pokorny, D. Vrba, O. Fiser, M. Salucci, and J. Vrba, “Systematic Optimization of Training and Setting of SVM-Based Microwave Stroke Classification: Numerical Simulations for 10 Port System,” IEEE J. Electromagn. RF Microw. Med. Biol., vol. 8, no. 3, pp. 273–281, Sep. 2024, doi:
1109/JERM.2024.3404119
L. de Vries, B. J. Emmer, C. B. Majoie, H. A. Marquering, and E. Gavves, “PerfU-Net: Baseline infarct estimation from CT perfusion source data for acute ischemic stroke,” Med. Image Anal., vol. 85, p. 102749, Apr. 2023, doi: doi.org/10.1016/j.media.2023.102749
M. Srikrishna, J. B. Pereira, R. A. Heckemann, F. Alshehri and G. Muhammad, "A few-shot learning-based ischemic stroke segmentation system using weighted MRI fusion," Image and Vision Computing, vol. 140, p. 104865, 2023. doi: 10.1016/j.imavis.2023.104865
A. Srinivas and J. P. Mosiganti, “A brain stroke detection model using soft voting based ensemble machine learning classifier,” Measurement: Sensors, vol. 29, p. 100871, Oct. 2023, doi: doi.org/10.1016/j.measen.2023.100871
O. Ozaltin, O. Coskun, O. Yeniay, and A. Subasi, “A Deep Learning Approach for Detecting Stroke from Brain CT Images Using OzNet,” Bioengineering, vol. 9, no. 12, p. 783, Dec. 2022, doi: 10.3390/bioengineering9120783
N. H. Ali, A. R. Abdullah, N. M. Saad, A. S. Muda, and E. E. M. Noor, “Automated Classification of Collateral Circulation for Ischemic Stroke in Cone-Beam CT Images Using VGG11: A Deep Learning Approach,” BioMedInformatics, vol. 4, no. 3, pp. 1692–1702, Jul. 2024, doi: doi.org/10.3390/biomedinformatics4030091
Y.-T. Chen, Y.-L. Chen, Y.-Y. Chen, Y.-T. Huang, H.-F. Wong, J.-L. Yan, and J.-J. Wang, “Deep Learning–Based Brain Computed Tomography Image Classification with Hyperparameter Optimization through Transfer Learning for Stroke,” Diagnostics, vol. 12, no. 4, p. 807, Mar. 2022, doi: 10.3390/diagnostics12040807
V. Mariano, J. A. Tobon Vasquez, M. R. Casu, and F. Vipiana, “Brain Stroke Classification via Machine Learning Algorithms Trained with a Linearized Scattering Operator,” Diagnostics, vol. 13, no. 1, p. 23, Dec. 2022, doi: doi.org/10.3390/diagnostics13010023
M. Uppal, D. Gupta, S. Juneja, T. R. Gadekallu, I. El Bayoumy, J. Hussain et al., “Enhancing accuracy in brain stroke detection: Multi-layer perceptron with Adadelta, RMSProp and AdaMax optimizers,” Front. Bioeng. Biotechnol., vol. 11, Sep. 2023, doi: doi.org/10.3389/fbioe.2023.1257591
F. Palsson, N. D. Forkert, L. Meyer, G. Broocks, F. Flottmann, M. E. Maros et al., “Prediction of tissue outcome in acute ischemic stroke based on single-phase CT angiography at admission,” Front. Neurol., vol. 15, Mar. 2024, doi: 10.3389/fneur.2024.1330497
M. A. Saleem, A. Javeed, W. Akarathanawat, A. Chutinet, N. C. Suwanwela, and W. Asdornwised, “Innovations in Stroke Identification: A Machine Learning-Based Diagnostic Model Using Neuroimages,” IEEE Access, vol. 12, pp. 35754–35764, 2024, doi: 10.1109/ACCESS.2024.3369673
M. M. Hossain, M. M. Ahmed, A. A. N. Nafi, M. R. Islam, M. S. Ali, and J. Haque, “A novel hybrid ViT-LSTM model with explainable AI for brain stroke detection and classification in CT images: A case study of Rajshahi region,” Comput. Biol. Med., vol. 186, p. 109711, Jan. 2025, doi: doi.org/10.1016/j.compbiomed.2025.109711
Ö. Polat, Z. Dokur, and T. Ölmez, “Classification of brain strokes using divergence-based convolutional neural networks,” Biomed. Signal Process. Control, vol. 93, p. 106193, Jul. 2024, doi: doi.org/10.1016/j.bspc.2024.106193
M. Soltanpour, P. Boulanger, and B. Buck, “CT perfusion map synthesis from CTP dynamic images using a learned LSTM generative adversarial network for acute ischemic stroke assessment,” J. Med. Syst., vol. 48, no. 1, p. 37, Apr. 2024, doi: 10.1007/s10916-024-02054-2
B. Kaya and M. Önal, “A CNN transfer learning-based approach for segmentation and classification of brain stroke from noncontrast CT images,” Int. J. Imaging Syst. Technol., vol. 33, no. 4, pp. 1335–1352, Jul. 2023, doi: 10.1002/ima.22864
M. A. Inamdar, A. Gudigar, U. Raghavendra, M. Salvi, R. R. A. B. R. Aman, and N. F. M. Gowdh, “A Dual-Stream Deep Learning Architecture With Adaptive Random Vector Functional Link for Multi-Center Ischemic Stroke Classification,” IEEE Access, vol. 13, pp. 46638–46658, 2025, doi: 10.1109/ACCESS.2025.3550344
A. Abumihsan, A. Yousef Owda, M. Owda, M. Abumohsen, L. Stergioulas, and M. Ahmad Abu Amer, “A Novel Hybrid Model for Brain Ischemic Stroke Detection Using Feature Fusion and Convolutional Block Attention Module,” IEEE Access, vol. 13, pp. 44466–44483, 2025, doi: 10.1109/ACCESS.2025.3549269
C. H. Patel, D. Undaviya, H. Dave, S. Degadwala, and D. Vyas, “EfficientNetB0 for brain stroke classification on computed tomography scan,” in Proc. 2nd Int. Conf. Appl. Artif. Intell. Comput. (ICAAIC), May 2023, pp. 713–718, doi: 10.1109/ICAAIC56838.2023.10141195
M. A. R. Afridi, “Brain stroke CT image dataset,” Kaggle, 2021. [Online]. Available: https://www.kaggle.com/datasets/afridirahman/brain-stroke-ct-image-dataset [Accessed: December 12, 2024].
M. A. R. Afridi, Z. Ali, R. Muhammad, A. Ahmad, and I. Alam, “Age and gender specific stroke risk factors in a teaching hospital in Khyber Pakhtunkhwa,” J. Postgraduate Med. Inst., vol. 29, no. 2, pp. 79–85, 2015.
I. Gulrajani, F. Ahmed, M. Arjovsky, V. Dumoulin, and A. C. Courville, “Improved training of Wasserstein GANs,” in Advances in Neural Information Processing Systems, vol. 30, 2017.
E. Aksan, M. Kaufmann, and O. Hilliges, “A Spatio-Temporal Transformer for 3D Human Motion Prediction,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), Montreal, QC, Canada, 2021, pp. 11884–11893. doi: 10.1109/3DV53792.2021.00066
Z. Zhao, S. Singh, H. Lee, Z. Zhang, A. Odena, and H. Zhang, “Improved consistency regularization for GANs,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 35, no. 12, pp. 11033–11041, 2021.
S. Deepak and P. M. Ameer, "Brain tumor categorization from imbalanced MRI dataset using weighted loss and deep feature fusion," Neurocomputing, vol. 520, pp. 94–102, 2023, doi: 10.1016/j.neucom.2022.11.039.
S. K. UmaMaheswaran, F. Ahmad, R. Hegde, A. M. Alwakeel, and S. R. Zahra, “Enhanced non-contrast computed tomography images for early acute stroke detection using machine learning approach,” Expert Systems with Applications, vol. 240, p. 122559, 2024, doi: doi.org/10.1016/j.eswa.2023.122559
M. A. Sabir and F. Ashraf, “Development of a novel deep convolutional neural network model for early detection of brain stroke using CT scan images,” Multimedia Tools and Applications, vol. 2024, pp. 1–25, Apr. 2024, doi: doi.org/10.1007/s11042-024-19001-5
M. Gupta, P. Meghana, K. H. Reddy, and P. Supraja, “Predicting Brain Stroke Using IoT-Enabled Deep Learning and Machine Learning: Advancing Sustainable Healthcare,” in Sustainable Development through Machine Learning, AI and IoT, P. Whig, N. Silva, A. A. Elngar, N. Aneja, and P. Sharma, Eds., Communications in Computer and Information Science, vol. 1939, 2023, doi: https://doi.org/10.1007/978-3-031-47055-4_10
R. Mena, A. Macas, E. Pelaez, F. Loayza, and H. Franco-Maldonado, “A Pipeline for Segmenting and Classifying Brain Lesions Caused by Stroke: A Machine Learning Approach,” in Information Systems and Technologies, A. Rocha, H. Adeli, G. Dzemyda, and F. Moreira, Eds., WorldCIST 2022, Lecture Notes in Networks and Systems, vol. 470, 2022, doi: doi.org/10.1007/978-3-031-04829-6_37
N. A. Chowdhury et al., “A novel approach to detect stroke from 2D images using deep learning,” in Proc. 2nd Int. Conf. Big Data, IoT and Mach. Learn. (BIM 2023), M. S. Arefin, M. S. Kaiser, T. Bhuiyan, N. Dey, and M. Mahmud, Eds., Lecture Notes in Networks and Systems, vol. 867, Singapore: Springer, 2024, doi: doi.org/10.1007/978-981-99-8937-9_17
Copyright (c) 2025 Fouzi Lezzar, Seif Eddine Mili
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution-ShareAlikel 4.0 International (CC BY-SA 4.0) that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
.png)
.png)
.png)
.png)
.png)
