Skip to main content
SpringerLink
Log in

A robust ischemic stroke lesion segmentation technique using two-pathway 3D deep neural network in MR images

  • Published:
Multimedia Tools and Applications Aims and scope Submit manuscript

This article has been updated

Abstract

Ischemic stroke is one of the major causes of disability and death of humans. It is a most common disease in aged people which may lead to long-term disability. So, accurate stroke lesion identification and quantification within a short period are the most important tasks in treatment planning. Generally, the supervised and semi-supervised based methods have succeeded in achieving promising performance for the segmentation of acute ischemic stroke lesions, however, few deep learning-based methods have been proposed in recent years successfully. In the present work, a robust deep neural network based on two-pathway 3D convolutional neural network has been proposed to identify the accurate boundary regions of the acute and sub-acute ischemic stroke lesions. The proposed two-pathway 3D CNN not only focuses on segmenting abnormal tissues on individual brain slices but also considers the information from preceding and succeeding slices to establish connectivity among the different slices. This approach allows the model to have a more comprehensive understanding of the brain structures and abnormalities being analyzed. The local pathway can capture fine-grained details and local patterns within each slice, enabling precise segmentation of aberrant tissues. Meanwhile, the contextual pathway considers the spatial dependencies and temporal information between slices, enhancing the model’s ability to detect and incorporate the connectivity between different brain regions. Generally, most of the existing deep learning-based methods utilized single MRI modalities (either DWI or FLAIR) to segment the acute ischemic stroke lesions because these two MRI modalities are the most sensitive to find out and quantifying the acute ischemic stroke lesions. In this present work, various MRI modalities have been utilized to improve the performance of the proposed model by accurately identifying the boundary regions of the acute and sub-acute ischemic stroke lesions. The current study explores the potential benefits of incorporating intensity normalization and data augmentation during the pre-processing stage to address the challenges associated with imbalanced stroke labels. The proposed model is tested with the ISLES2015 datasets and obtains promising results as compared to the existing deep learning-based methods depending on various metrics such as dice similarity coefficient (DSC), sensitivity, and positive predictive value (PPV). Furthermore, a significant gain is achieved around the boundaries of the sub-regions of the stroke lesions.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Fig. 12
Fig. 13
Fig. 14
Fig. 15
Fig. 16
Fig. 17
Fig. 18
Fig. 19
Fig. 20
Fig. 21
Fig. 22

Similar content being viewed by others

Change history

  • 23 October 2023

    On page 34 of the PDF version, table 8 was moved before figure 22.

References

  1. Clèrigues A, Valverde S, Bernal J, Freixenet J, Oliver A, Lladó X (2020) Acute and sub-acute stroke lesion segmentation from multimodal mri. Comput Methods Programs Biomed 194:105521

    Article  Google Scholar 

  2. Chen L, Bentley P, Rueckert D (2017) Fully automatic acute ischemic lesion segmentation in dwi using convolutional neural networks. NeuroImage Clin 15:633–643

    Article  Google Scholar 

  3. Zhao B, Liu Z, Liu G, Cao C, Jin S, Wu H, Ding S (2021) Deep learning-based acute ischemic stroke lesion segmentation method on multimodal mr images using a few fully labeled subjects. Comput Math Methods Med

  4. Zhang L, Song R, Wang Y, Zhu C, Liu J, Yang J, Liu L (2020) Ischemic stroke lesion segmentation using multi-plane information fusion. IEEE Access 8:45715–45725

    Article  Google Scholar 

  5. Huang B, Tan G, Dou H, Cui Z, Song Y, Zhou T (2022) Mutual gain adaptive network for segmenting brain stroke lesions. Appl Soft Comput 129:109568

    Article  Google Scholar 

  6. Yalçın S, Vural H (2022) Brain stroke classification and segmentation using encoder-decoder based deep convolutional neural networks. Comput Biol Med 149:105941

    Article  Google Scholar 

  7. Tursynova A, Omarov B (2021) 3d u-net for brain stroke lesion segmentation on isles 2018 dataset. In: 2021 16th international conference on electronics computer and computation (ICECCO). IEEE pp 1–4

  8. Zhang Y, Liu S, Li C, Wang J (2022) Application of deep learning method on ischemic stroke lesion segmentation. J Shanghai Jiaotong Univ (Sci) 1–13

  9. Kumar A (2023) Study and analysis of different segmentation methods for brain tumor mri application. Multimed Tools Appl 82(5):7117–7139

    Article  Google Scholar 

  10. Kumar A, Chauda P, Devrari A (2021) Machine learning approach for brain tumor detection and segmentation. Int J Org Coll Intell 11(3):68–84

    Google Scholar 

  11. Goel A, Goel AK, Kumar A (2022) The role of artificial neural network and machine learning in utilizing spatial information. Spat Inf Res 1–11

  12. Bal A, Banerjee M, Sharma P, Maitra M (2020) Gray matter segmentation and delineation from positron emission tomography (pet) image. In: Emerging technology in modelling and graphics. Springer, pp 359–372

  13. Tomita N, Jiang S, Maeder ME, Hassanpour S (2020) Automatic post-stroke lesion segmentation on mr images using 3d residual convolutional neural network. NeuroImage Clin 27:102276

    Article  Google Scholar 

  14. Clerigues A, Valverde S, Bernal J, Freixenet J, Oliver A, Lladó X (2019) Acute ischemic stroke lesion core segmentation in ct perfusion images using fully convolutional neural networks. Comput Biol Med 115:103487

    Article  Google Scholar 

  15. Zhao B, Ding S, Wu H, Liu G, Cao C, Jin S, Liu Z Automatic acute ischemic stroke lesion segmentation using semi-supervised learning. arXiv:1908.03735

  16. Noh H, Hong S, Han B (2015) Learning deconvolution network for semantic segmentation. In: Proceedings of the IEEE international conference on computer vision. pp 1520–1528

  17. Bal A, Banerjee M, Chaki R, Sharma P (2020) An efficient method for pet image denoising by combining multi-scale transform and non-local means. Multime Tools Appl 79:29087–29120

    Article  Google Scholar 

  18. Bal A, Banerjee M, Sharma P, Maitra M (2019) An efficient wavelet and curvelet-based pet image denoising technique. Med Biol Eng Comput 57:2567–2598

    Article  Google Scholar 

  19. Li C, Gore JC, Davatzikos C (2014) Multiplicative intrinsic component optimization (mico) for mri bias field estimation and tissue segmentation. Magn Reson Imaging 32(7):913–923

    Article  Google Scholar 

  20. Nyúl LG, Udupa JK, Zhang X (2000) New variants of a method of mri scale standardization. IEEE Trans Med Imaging 19(2):143–150

    Article  Google Scholar 

  21. Kamnitsas K, Ledig C, Newcombe VF, Simpson JP, Kane AD, Menon DK, Rueckert D, Glocker B (2017) Efficient multi-scale 3d cnn with fully connected crf for accurate brain lesion segmentation. Med Image Anal 36:61–78

    Article  Google Scholar 

  22. Bal A, Banerjee M, Chakrabarti A, Sharma P (2022) Mri brain tumor segmentation and analysis using rough-fuzzy c-means and shape based properties. J King Saud Univ Comput Inf Sci 34(2):115–133

    Google Scholar 

  23. Bal A, Banerjee M, Chaki R, Sharma P (2021) An efficient brain tumor image classifier by combining multi-pathway cascaded deep neural network and handcrafted features in mr images. Med Biol Eng Comput 59(7–8):1495–1527

    Article  Google Scholar 

  24. Bal A, Banerjee M, Sharma P, Maitra M (2018) Brain tumor segmentation on mr image using k-means and fuzzy-possibilistic clustering. In: 2018 2nd international conference on electronics, materials engineering & Nano-Technology (IEMENTech). IEEE, pp 1–8

  25. Bal A, Banerjee M, Sharma P, Chaki R (2020) A multi-class image classifier for assisting in tumor detection of brain using deep convolutional neural network. In: Advanced computing and systems for security. Springer, pp 93–111

  26. Pereira S, Pinto A, Alves V, Silva CA (2016) Brain tumor segmentation using convolutional neural networks in mri images. IEEE Trans Med Imaging 35(5):1240–1251

    Article  Google Scholar 

  27. Khan H, Shah PM, Shah MA, Ul Islam S, Rodrigues JJ Cascading handcrafted features and convolutional neural network for iot-enabled brain tumor segmentation. Comput Commun

  28. Wang G, Li W, Vercauteren T, Ourselin S (2019) Automatic brain tumor segmentation based on cascaded convolutional neural networks with uncertainty estimation. Front Comput Neurosci 13:56

    Article  Google Scholar 

  29. Wong SC, Gatt A, Stamatescu V, McDonnell MD (2016) Understanding data augmentation for classification: when to warp? In: Digital image computing: techniques and applications (DICTA), 2016 International conference on. IEEE, pp 1–6

  30. Fawzi A, Samulowitz H, Turaga D, Frossard P (2016) (2016) Adaptive data augmentation for image classification. Image processing (ICIP). IEEE international conference on, Ieee, pp 3688–3692

    Google Scholar 

  31. Wang J, Perez L (2017) The effectiveness of data augmentation in image classification using deep learning. Tech. Rep, Technical report

    Google Scholar 

  32. Glorot X, Bengio Y (2010) Understanding the difficulty of training deep feedforward neural networks. In: Proceedings of the thirteenth international conference on artificial intelligence and statistics. pp 249–256

  33. Srivastava N, Hinton G, Krizhevsky A, Sutskever I, Salakhutdinov R (2014) Dropout: a simple way to prevent neural networks from overfitting. J Mach Learn Res 15(1):1929–1958

    MathSciNet  Google Scholar 

  34. Maas AL, Hannun AY, Ng AY (2013) Rectifier nonlinearities improve neural network acoustic models. In: Proc icml, vol 30. p 3

  35. Havaei M, Davy A, Warde-Farley D, Biard A, Courville A, Bengio Y, Pal C, Jodoin P-M, Larochelle H (2017) Brain tumor segmentation with deep neural networks. Med Image Anal 35:18–31

    Article  Google Scholar 

  36. Hussain S, Anwar SM, Majid M (2017) Brain tumor segmentation using cascaded deep convolutional neural network. In: Engineering in medicine and biology society (EMBC), 2017 39th annual international conference of the IEEE. IEEE, pp 1998–2001

  37. Zhao L, Jia K (2016) Multiscale cnns for brain tumor segmentation and diagnosis. Comput Math Methods Med

  38. Gonzalez RC (2009) Digital image processing. Pearson Education India

  39. Maier O, Menze BH, von der Gablentz J, Häni L, Heinrich MP, Liebrand M, Winzeck S, Basit A, Bentley P, Chen L et al (2017) Isles 2015-a public evaluation benchmark for ischemic stroke lesion segmentation from multispectral mri. Med Image Anal 35:250–269

    Article  Google Scholar 

Download references

Acknowledgements

The current work was carried out with inspirational support from the Board of Research in Nuclear Sciences (ref no. 34/14/13/2016-BRNS/34044). The heartiest thanks to Dr. Punit Sharma, for providing precious suggestions and validating the results as a neuro-consultant at Apollo Gleneagles Hospital, Kolkata, India.

Author information

Authors and Affiliations

  1. A.K. Choudhury School of Information Technology University of Calcutta, Kolkata, India

    Abhishek Bal & Rituparna Chaki

  2. RCC Institute of Information Technology, Kolkata, India

    Minakshi Banerjee

  3. Apollo Gleneagles Hospital, Kolkata, India

    Punit Sharma

Authors
  1. Abhishek Bal
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Minakshi Banerjee
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Rituparna Chaki
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Punit Sharma
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Abhishek Bal.

Ethics declarations

Conflicts of interest

The authors have declared that there is no conflict of interest regarding the publication of this article.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

On page 34 of the PDF version, table 8 was moved before figure 22.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Bal, A., Banerjee, M., Chaki, R. et al. A robust ischemic stroke lesion segmentation technique using two-pathway 3D deep neural network in MR images. Multimed Tools Appl 83, 41485–41524 (2024). https://doi.org/10.1007/s11042-023-16689-9

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11042-023-16689-9

Keywords

Search

Navigation