U-Net-Based Models for Precise Brain Stroke Segmentation

Year 2025, Volume: 7 Issue: 1, 50 - 60

Suat İnce , Ismail Kunduracioglu , Bilal Bayram , Ishak Pacal

https://doi.org/10.51537/chaos.1605529

Abstract

Ischemic stroke, a widespread neurological condition with a substantial mortality rate, necessitates accurate delineation of affected regions to enable proper evaluation of patient outcomes. However, such precision is complicated by factors like variable lesion sizes, noise interference, and the overlapping intensity characteristics of different tissue structures. This research addresses these issues by focusing on the segmentation of Diffusion Weighted Imaging (DWI) scans from the ISLES 2022 dataset and conducting a comparative assessment of three advanced deep learning models: the U-Net framework, its U-Net++ extension, and the Attention U-Net. Applying consistent evaluation criteria specifically, Intersection over Union (IoU), Dice Similarity Coefficient (DSC), and recall the Attention U-Net emerged as the superior choice, establishing record high values for IoU (0.8223) and DSC (0.9021). Although U-Net achieved commendable recall, its performance lagged behind that of U-Net++ in other critical measures. These findings underscore the value of integrating attention mechanisms to achieve more precise segmentation. Moreover, they highlight that the Attention U-Net model is a reliable candidate for medical imaging tasks where both accuracy and efficiency hold paramount importance, while U Net and U Net++ may still prove suitable in certain niche scenarios.

Keywords

Medical image analysis, Ishemic stroke, Brain stroke, Segmentation, U-Net, Deep learning

References

• Abdmouleh, N., A. Echtioui, F. Kallel, and A. B. Hamida, 2022 Modified u-net architeture based ischemic stroke lesions segmentation. In 2022 IEEE 21st International Conference on Sciences and Techniques of Automatic Control and Computer Engineering (STA), pp. 361–365.
• Alkan, T., Y. Dokuz, A. Ecemi¸s, A. Bozda˘ g, and S. S. Durduran, 2023 Using machine learning algorithms for predicting real estate values in tourism centers. Soft Computing 27: 2601–2613.
• Alshawi, R., M. T. Hoque, M. M. Ferdaus, M. Abdelguerfi, K. Niles, et al., 2023 Dual attention u-net with feature infusion: Pushing the boundaries of multiclass defect segmentation. Unpublished .
• Ansari, M. Y., Y. Yang, S. Balakrishnan, J. Abinahed, A. Al-Ansari, et al., 2022 A lightweight neural network with multiscale feature enhancement for liver ct segmentation. Scientific Reports 12: 14153.
• Ashburner, J. and K. J. Friston, 2005 Unified segmentation. NeuroImage 26: 839–851.
• Aslan, E., 2024 LSTM-ESA Hibrit Modeli ile MR Goruntulerinden Beyin Tumorunun Siniflandirilmasi. Adiyaman Universitesi Muhendislik Bilimleri Dergisi 11: 63–81.
• Aslan, E. and Y. Ozupak, 2025 Detection of road extraction from satellite images with deep learning method. Cluster Computing 28: 72.
• Bal, A., M. Banerjee, P. Sharma, and M. Maitra, 2019 An efficient wavelet and curvelet-based pet image denoising technique. Medical & Biological Engineering & Computing 57: 2567–2598.
• Bayram, B., I. Kunduracioglu, S. Ince, and I. Pacal, 2025 A systematic review of deep learning in mri-based cerebral vascular occlusion-based brain diseases. Neuroscience .
• Burukanli, M. and N. Yumu¸sak, 2024 Tfradmcov: a robust transformer encoder based model with adam optimizer algorithm for covid-19 mutation prediction. Connection Science 36: 2365334.
• Çiçek, Ö., A. Abdulkadir, S. S. Lienkamp, T. Brox, and O. Ronneberger, 2016 3d u-net: Learning dense volumetric segmentation from sparse annotation. In Proceedings of the International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI), pp. 424–432.
• Celik, M., A. S. Dokuz, A. Ecemis, and E. Erdogmus, 2025 Discovering and ranking urban social clusters out of streaming social media datasets. Concurrency and Computation: Practice and Experience 37: e8314.
• Chen, G., Z. Li, J.Wang, J.Wang, S. Du, et al., 2023 An improved 3d kiu-net for segmentation of liver tumor. Computers in Biology and Medicine 160: 107006.
• Chen, J., Y. Lu, Q. Yu, X. Luo, E. Adeli, et al., 2021 Transunet: Transformers make strong encoders for medical image segmentation. Unpublished .
• Chen, L., P. Bentley, and D. Rueckert, 2017 Fully automatic acute ischemic lesion segmentation in dwi using convolutional neural networks. NeuroImage: Clinical 15: 633–643.
• Clèrigues, A., S. Valverde, J. Bernal, J. Freixenet, A. Oliver, et al., 2020 Acute and sub-acute stroke lesion segmentation from multimodal mri. Computer Methods and Programs in Biomedicine 194: 105521.
• Dice, L., 1945 Measures of the amount of ecologic homeostasis. Science 113: 297–302.
• Ding, Y., W. Zheng, J. Geng, Z. Qin, K.-K. R. Choo, et al., 2022 Mvfusfra: A multi-view dynamic fusion framework for multimodal brain tumor segmentation. IEEE Journal of Biomedical and Health Informatics 26: 1570–1581.
• Dosovitskiy, A., L. Beyer, A. Kolesnikov, D.Weissenborn, X. Zhai, et al., 2020 An image is worth 16x16 words: Transformers for image recognition at scale. arXiv preprint arXiv:2010.11929 .
• Edlow, B. L., S. Hurwitz, and J. A. Edlow, 2017 Diagnosis of dwinegative acute ischemic stroke. Neurology 89: 256–262.
• Everingham, M. and et al., 2010 The pascal visual object classes (voc) challenge. International Journal of Computer Vision 88: 303–338.
• Goel, A., A. K. Goel, and A. Kumar, 2023 The role of artificial neural network and machine learning in utilizing spatial information. Spatial Information Research 31: 275–285.
• Hernandez Petzsche, M. R., E. de la Rosa, U. Hanning, R. Wiest, W. Valenzuela, et al., 2022 Isles 2022: A multi-center magnetic resonance imaging stroke lesion segmentation dataset. Scientific Data 9: 762.
• Hossain, M. S., J. M. Betts, and A. P. Paplinski, 2021 Dual focal loss to address class imbalance in semantic segmentation. Neurocomputing 462: 69–87.
• Huang, B., G. Tan, H. Dou, Z. Cui, Y. Song, et al., 2022 Mutual gain adaptive network for segmenting brain stroke lesions. Applied Soft Computing 129: 109568.
• Jauch, E. C., J. L. Saver, H. P. Adams, A. Bruno, J. J. B. Connors, et al., 2013 Guidelines for the early management of patients with acute ischemic stroke. Stroke 44: 870–947.
• Johnson, L., R. Newman-Norlund, A. Teghipco, C. Rorden, L. Bonilha, et al., 2024 Progressive lesion necrosis is related to increasing aphasia severity in chronic stroke. NeuroImage: Clinical 41: 103566.
• Kamnitsas, K., C. Ledig, V. F. J. Newcombe, J. P. Simpson, A. D. Kane, et al., 2017 Efficient multi-scale 3d cnn with fully connected crf for accurate brain lesion segmentation. Medical Image Analysis 36: 61–78.
• Karani, N., E. Erdil, K. Chaitanya, and E. Konukoglu, 2021 Testtime adaptable neural networks for robust medical image segmentation. Medical Image Analysis 68: 101907.
• Kench, S. and S. J. Cooper, 2021 Generating 3d structures from a 2d slice with gan-based dimensionality expansion. Nature Machine Intelligence .
• Kilicarslan, S. and I. Pacal, 2023 Domates yapraklarıinda hastalık tespiti için transfer ogrenme metotlarınn kullanılması. Mühendislik Bilimleri ve Ara¸stırmaları Dergisi 5: 215–222.
• Kim, Y.-C., J.-E. Lee, I. Yu, H.-N. Song, I.-Y. Baek, et al., 2019 Evaluation of diffusion lesion volume measurements in acute ischemic stroke using encoder-decoder convolutional network. Stroke 50: 1444–1451.
• Kumar, A., P. Chauda, and A. Devrari, 2021 Machine learning approach for brain tumor detection and segmentation. International Journal of Organizational and Collective Intelligence 11: 68–84.
• Kunduracioglu, I., 2024a Cnn models approaches for robust classification of apple diseases. Computer and Decision Making: An International Journal 1: 235–251.
• Kunduracioglu, I., 2024b Utilizing resnet architectures for identification of tomato diseases. Journal of Intelligent Decision Making and Information Science 1: 104–119.
• Kunduracioglu, I. and I. Pacal, 2024 Advancements in deep learning for accurate classification of grape leaves and diagnosis of grape diseases. Journal of Plant Diseases and Protection .
• Lee, K.-Y., C.-C. Liu, D. Y.-T. Chen, C.-L.Weng, H.-W. Chiu, et al., 2023 Automatic detection and vascular territory classification of hyperacute staged ischemic stroke on diffusion weighted image using convolutional neural networks. Scientific Reports 13: 404.
• Li, T., X. An, Y. Di, C. Gui, Y. Yan, et al., 2024 Srsnet: Accurate segmentation of stroke lesions by a two-stage segmentation framework with asymmetry information. Expert Systems with Applications 254: 124329.
• Li, Z., D. Li, C. Xu, W. Wang, Q. Hong, et al., 2022 Tfcns: A cnntransformer hybrid network for medical image segmentation. In Proceedings of the International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI), pp. 781– 792.
• Liu, Y., W. Cui, Q. Ha, X. Xiong, X. Zeng, et al., 2021 Knowledge transfer between brain lesion segmentation tasks with increased model capacity. Computerized Medical Imaging and Graphics 88: 101842.
• Maier, O., B. H. Menze, J. von der Gablentz, L. Häni, M. P. Heinrich, et al., 2017 Isles 2015 - a public evaluation benchmark for ischemic stroke lesion segmentation from multispectral mri. Medical Image Analysis 35: 250–269.
• Moon, H. S., L. Heffron, A. Mahzarnia, B. Obeng-Gyasi, M. Holbrook, et al., 2022 Automated multimodal segmentation of acute ischemic stroke lesions on clinical mr images. Magnetic Resonance Imaging 92: 45–57.
• Nielsen, A., M. B. Hansen, A. Tietze, and K. Mouridsen, 2018 Prediction of tissue outcome and assessment of treatment effect in acute ischemic stroke using deep learning. Stroke 49: 1394– 1401.
• Oktay, O., J. Schlemper, L. L. Folgoc, M. Lee, M. Heinrich, et al., 2018 Attention u-net: Learning where to look for the pancreas. Medical Image Analysis 53: 197–207.
• Ozdemir, B. and I. Pacal, 2025 An innovative deep learning framework for skin cancer detection employing convnextv2 and focal self-attention mechanisms. Results in Engineering 25: 103692.
• Pacal, I., 2025 Investigating deep learning approaches for cervical cancer diagnosis: a focus on modern image-based models. European Journal of Gynaecological Oncology 46: 125–141.
• Pacal, I., I. Kunduracioglu, M. H. Alma, M. Deveci, S. Kadry, et al., 2024 A systematic review of deep learning techniques for plant diseases. Artificial Intelligence Review 57: 304.
• Paçal, I. and I. Kunduracıo˘ glu, 2024 Data-efficient vision transformer models for robust classification of sugarcane. Journal of Soft Computing and Decision Analytics 2: 258–271.
• Ronneberger, O., P. Fischer, and T. Brox, 2015 U-net: Convolutional networks for biomedical image segmentation. In Proceedings of the International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI), pp. 234–241.
• Roth, G. A., D. Abate, K. H. Abate, S. M. Abay, C. Abbafati, et al., 2018 Global, regional, and national age-sex-specific mortality for 282 causes of death in 195 countries and territories, 1980-2017: a systematic analysis for the global burden of disease study 2017. The Lancet 392: 1736–1788.
• Sacco, R. L., S. E. Kasner, J. P. Broderick, L. R. Caplan, J. J. B. Connors, et al., 2013 An updated definition of stroke for the 21st century. Stroke 44: 2064–2089.
• Salvi, M., U. R. Acharya, F. Molinari, and K. M. Meiburger, 2021 The impact of pre- and post-image processing techniques on deep learning frameworks: A comprehensive review for digital pathology image analysis. Computers in Biology and Medicine 128: 104129.
• Sarvamangala, D. R. and R. V. Kulkarni, 2022 Convolutional neural networks in medical image understanding: a survey. Evolutionary Intelligence 15: 1–22.
• Saver, J. L., 2006 Time is brainâ˘Aˇ Tquantified. Stroke 37: 263–266.
• Schlemper, J., O. Oktay, M. Schaap, M. Heinrich, B. Kainz, et al., 2019 Attention gated networks: Learning to leverage salient regions in medical images. Medical Image Analysis 53: 197–207.
• The GBD, . L. R. O. S. C., 2018 Global, regional, and country-specific lifetime risks of stroke, 1990 and 2016. New England Journal of Medicine 379: 2429–2437.
• Tomita, N., S. Jiang, M. E. Maeder, and S. Hassanpour, 2020 Automatic post-stroke lesion segmentation on mr images using 3d residual convolutional neural network. NeuroImage: Clinical 27: 102276.
• Tursynova, A. and B. Omarov, 2021 3d u-net for brain stroke lesion segmentation on isles 2018 dataset. In 2021 16th International Conference on Electronics Computer and Computation (ICECCO), pp. 1–4.
• van Rijsbergen, C. J., 1979 Information Retrieval. Butterworth. Verclytte, S., R. Gnanih, S. Verdun, T. Feiweier, B. Clifford, et al., 2023 Ultrafast mri using deep learning echoplanar imaging for a comprehensive assessment of acute ischemic stroke. European Radiology 33: 3715–3725.
• Wang, G., T. Song, Q. Dong, M. Cui, N. Huang, et al., 2020 Automatic ischemic stroke lesion segmentation from computed tomography perfusion images by image synthesis and attentionbased deep neural networks. Medical Image Analysis 65: 101787.
• Wang, Z., B. Wang, C. Zhang, and Y. Liu, 2023 Defense against adversarial patch attacks for aerial image semantic segmentation by robust feature extraction. Remote Sensing 15: 1690.
• Wong, K. K., J. S. Cummock, G. Li, R. Ghosh, P. Xu, et al., 2022 Automatic segmentation in acute ischemic stroke: Prognostic significance of topological stroke volumes on stroke outcome. Stroke 53: 2896–2905.
• Woo, S., J. Park, J.-Y. Lee, and I. S. Kweon, 2018 Cbam: Convolutional block attention module. In Proceedings of the European Conference on Computer Vision (ECCV), pp. 3–19.
• Wu, Z., X. Zhang, F. Li, S. Wang, L. Huang, et al., 2023 W-net: A boundary-enhanced segmentation network for stroke lesions. Expert Systems with Applications 230: 120637.
• Wu, Z., X. Zhang, F. Li, S.Wang, and J. Li, 2024 A feature-enhanced network for stroke lesion segmentation from brain mri images. Computers in Biology and Medicine 174: 108326.
• Xiao, X., S. Lian, Z. Luo, and S. Li, 2018 Weighted res-unet for high-quality retina vessel segmentation. In 2018 9th International Conference on Information Technology in Medicine and Education (ITME), pp. 327–331.
• Xie, Y., J. Zhang, C. Shen, and Y. Xia, 2021 Cotr: Efficiently bridging cnn and transformer for 3d medical image segmentation. In Proceedings of the International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI), pp. 171–180.
• Yalçın, S. and H. Vural, 2022 Brain stroke classification and segmentation using encoder-decoder based deep convolutional neural networks. Computers in Biology and Medicine 149: 105941.
• Yang, H., W. Huang, K. Qi, C. Li, X. Liu, et al., 2019 Clci-net: Cross-level fusion and context inference networks for lesion segmentation of chronic stroke. In Proceedings of the International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI), pp. 266–274.
• Yuan, F., Z. Zhang, and Z. Fang, 2023 An effective cnn and transformer complementary network for medical image segmentation. Pattern Recognition 136: 109228.
• Zhang, L., R. Song, Y. Wang, C. Zhu, J. Liu, et al., 2020 Ischemic stroke lesion segmentation using multi-plane information fusion. IEEE Access 8: 45715–45725.
• Zhang, Y. Q., A. F. Liu, F. Y. Man, Y. Y. Zhang, C. Li, et al., 2022 Mri radiomic features-based machine learning approach to classify ischemic stroke onset time. Journal of Neurology pp. 1–11.
• Zhao, B., S. Ding, H. Wu, G. Liu, C. Cao, et al., 2019 Automatic acute ischemic stroke lesion segmentation using semisupervised learning. Neurocomputing .
• Zhou, Z., M. M. Rahman Siddiquee, N. Tajbakhsh, and J. Liang, 2018 Unet++: A nested u-net architecture for medical image segmentation. In Proceedings of the European Conference on Computer Vision (ECCV), pp. 3–11.
• Zhou, Z., M. M. R. Siddiquee, N. Tajbakhsh, and J. Liang, 2020 Unet++: Redesigning skip connections to exploit multiscale features in image segmentation. IEEE Transactions on Medical Imaging 39: 1856–1867.
• Zhuang, X. and J. Shen, 2016 Multi-scale patch and multi-modality atlases for whole heart segmentation of mri. Medical Image Analysis 31: 77–87.