Beyin Bilgisayarlı Tomografi Görüntülerinde Derin Öğrenme Tabanlı İskemik İnme Hastalığı Segmentasyonu

Abstract

İnme, beyindeki işlevlerin doğru şekilde yerine getirilmesini engelleyen ve kan akışı eksikliği (iskemik) ya da kanama (hemorajik) gibi nedenlerle ortaya çıkan beyin hücre ölümüdür. İskemik inme, kan akışının beyin bölgelerine akmasını önleyen serebrovasküler sistemdeki bir tıkanıklık nedeniyle ortaya çıkan yaygın bir inme türüdür. İnme değerlendirmesinde sıklıkla Bilgisayarlı Tomografi (BT) taraması kullanılmaktadır ve BT görüntüleriyle iskemik inmenin hızlı ve doğru teşhisi, uygun tedavinin belirlenmesi için kritik öneme sahiptir. Uzmanların yoğun programları ve sağlık tesislerine başvuran çok sayıda hastanın olması gibi çeşitli faktörler nedeniyle iskemik inmenin manuel teşhisi hataya açık olabilmektedir. Bu nedenle, bu makalede, BT görüntüleri üzerinden segmentasyon yoluyla iskemik inmeyi otomatik olarak teşhis etmek için derin öğrenme tabanlı bir arayüz geliştirilmiş; bu sayede uzmanların teşhis süresi ve iş yükünün azaltılması hedeflenmiştir. Iskemik inme segmentasyonunda otomatik özellik çıkarımını sağlayan Evrişimli Sinir Ağları (CNN'ler), BT görüntülerindeki hastalıklı bölgeleri işaretlemek için kullanılmıştır. U-Net, U-Net VGG16, U-Net VGG19, Attention U-Net ve ResU-Net gibi CNN tabanlı mimariler, iskemik inme hastalığı segmentasyonunu karşılaştırmak için kullanılmıştır. ResU-Net, segmentasyon performansını daha da artırmak için mimarinin son katmanından sonra bir genişletme evrişim katmanı eklenerek modifiye edilmiştir. Ek olarak, iskemik inme hastalığı bölgesi için gerçek referans değerleri de içeren veri setindeki görüntü sayısını artırmak için veri artırma işlemi gerçekleştirilmiştir. Deneysel sonuçlara dayanarak, genişletme evrişimli olarak modifiye edilmiş ResU-Net, zar benzerlik katsayısı (DSC) ve Jaccard benzerlik katsayısı (IoU) açısından sırasıyla 98,45 % ve 96,95 % ile en yüksek performansı sağlamıştır. Deneysel sonuçlar, modifiye edilmiş ResU-Net mimarisinin iskemik inme hastalığı segmentasyonu için modern yaklaşımlardan daha iyi performans sergilediğini göstermektedir. Ayrıca modifiye edilmiş mimari, iskemik inme bölgesini segmente ederek hastalığın teşhisinde uzmanlara destek sağlayabilen yeni bir masaüstü uygulaması olan BrainSeg'e entegre edilmiştir.

Keywords

• Yapay Zekâ
• Derin Öğrenme
• İskemik İnme Hastalığı
• Evrişimsel Sinir Ağı

Deep Learning-Based Ischemic Stroke Segmentation on Brain Computed Tomography Images

Abstract

Stroke is brain cell death because of either lack of blood flow (ischemic) or bleeding (hemorrhagic) that prevents the brain from functioning properly in both conditions. Ischemic stroke is a common type of stroke caused by a blockage in the cerebrovascular system that prevents blood from flowing to brain regions and directly blocks blood vessels. Computed tomography (CT) scanning is frequently used in the evaluation of stroke, and rapid and accurate diagnosis of ischemic stroke with CT images is critical for determining the appropriate treatment. The manual diagnosis of ischemic stroke can be error-prone due to several factors, such as the busy schedules of specialists and the large number of patients admitted to healthcare facilities. Therefore, in this paper, a deep learning-based interface was developed to automatically diagnose the ischemic stroke through segmentation on CT images leading to a reduction on the diagnosis time and workload of specialists. Convolutional Neural Networks (CNNs) allow automatic feature extraction in ischemic stroke segmentation, utilized to mark the disease regions from CT images. CNN-based architectures, such as U-Net, U-Net VGG16, U-Net VGG19, Attention U-Net, and ResU-Net, were used to benchmark the ischemic stroke disease segmentation. To further improve the segmentation performance, ResU-Net was modified, adding a dilation convolution layer after the last layer of the architecture. In addition, data augmentation was performed to increase the number of images in the dataset, including the ground truths for the ischemic stroke disease region. Based on the experimental results, our modified ResU-Net with a dilation convolution provides the highest performance for ischemic stroke segmentation in dice similarity coefficient (DSC) and intersection over union (IoU) with 98.45 % and 96.95 %, respectively. The experimental results show that our modified ResU-Net outperforms the state-of-the-art approaches for ischemic stroke disease segmentation. Moreover, the modified architecture has been deployed into a new desktop application called BrainSeg, which can support specialists during the diagnosis of the disease by segmenting ischemic stroke.

Keywords

• Artificial Intelligence
• Deep Learning
• Ischemic Stroke Disease
• Convolutional Neural Network

References

1. Abdulkareem, K. H., Mohammed, M. A., Salim, A., Arif, M., Geman, O., Gupta, D., & Khanna, A. (2021). Realizing an effective COVID-19 diagnosis system based on machine learning and IOT in smart hospital environment. IEEE Internet of things journal, 8(21), 15919-15928.
2. Agrali, M., Soydemir, M. U., Gökçen, A., & Sahin, S. (2021). Deep Reinforcement Learning Based Controller Design for Model of The Vertical Take off and Landing System. Avrupa Bilim ve Teknoloji Dergisi, 26, 358-363.
3. Ağralı, M., Kilic, V., Onan, A., Koç, E. M., Koç, A. M., Büyüktoka, R. E., . . . Adıbelli, Z. (2023). DeepChestNet: Artificial intelligence approach for COVID-19 detection on computed tomography images. International Journal of Imaging Systems and Technology, 1-13.
4. Akosman, Ş. A., Öktem, M., Moral, Ö. T., & Kılıç, V. (2021). Deep Learning-based Semantic Segmentation for Crack Detection on Marbles. Paper presented at the 2021 29th Signal Processing and Communications Applications Conference (SIU).
5. Aljohani, A., & Alharbe, N. (2022). Generating Synthetic Images for Healthcare with Novel Deep Pix2Pix GAN. Electronics, 11(21), 3470.
6. Aydın, S., Çaylı, Ö., Kılıç, V., & Onan, A. (2022). Sequence-to-sequence video captioning with residual connected gated recurrent units. Avrupa Bilim ve Teknoloji Dergisi, 35, 380-386.
7. Castiglioni, I., Rundo, L., Codari, M., Di Leo, G., Salvatore, C., Interlenghi, M., . . . Sardanelli, F. (2021). AI applications to medical images: From machine learning to deep learning. Physica Medica, 83, 9-24.
8. Çaylı, Ö., Kılıç, V., Onan, A., & Wang, W. (2022). Auxiliary Classifier based Residual RNN for Image Captioning. Paper presented at the 2022 30th European Signal Processing Conference (EUSIPCO).

9. Das, S., Bhat, A. P., & Gogate, P. R. (2021). Degradation of dyes using hydrodynamic cavitation: Process overview and cost estimation. Journal of Water Process Engineering, 42, 102126.
10. Dina, A. S., Siddique, A., & Manivannan, D. (2023). A deep learning approach for intrusion detection in Internet of Things using focal loss function. Internet of Things, 100699.
11. Doǧan, V., Isık, T., Kılıç, V., & Horzum, N. (2022). A field-deployable water quality monitoring with machine learning-based smartphone colorimetry. Analytical Methods, 14(35), 3458-3466.
12. Doğan, V., & Kılıç, V. (2021). Akıllı Telefon Kullanarak Yapay Zeka Tabanlı Farenjit Tespiti: Artificial Intelligence Based Pharyngitis Detection Using Smartphone. Sağlık Bilimlerinde Yapay Zeka Dergisi ISSN, 1(2), 14-19.
13. Doğan, V., Yüzer, E., Kılıç, V., & Şen, M. (2021). Non-enzymatic colorimetric detection of hydrogen peroxide using a μPAD coupled with a machine learning-based smartphone app. Analyst, 146(23), 7336-7344.
14. Fetiler, B., Caylı, Ö., Moral, Ö. T., Kılıc, V., & Onan, A. (2021). Video captioning based on multi-layer gated recurrent unit for smartphones. Avrupa Bilim ve Teknoloji Dergisi, 32, 221-226.
15. Gölcez, T., Kilic, V., & Şen, M. (2021). A portable smartphone-based platform with an offline image-processing tool for the rapid paper-based colorimetric detection of glucose in artificial saliva. Analytical Sciences, 37(4), 561-567.
16. Gölcez, T., Kiliç, V., & Şen, M. (2019). Integration of a Smartphone Application with a ¼PAD for Rapid Colorimetric Detection of Glucose. Paper presented at the 2019 Medical Technologies Congress (TIPTEKNO).
17. Hui, H., Zhang, X., Li, F., Mei, X., & Guo, Y. (2020). A partitioning-stacking prediction fusion network based on an improved attention U-Net for stroke lesion segmentation. IEEE Access, 8, 47419-47432.
18. Karthik, R., Menaka, R., Johnson, A., & Anand, S. (2020). Neuroimaging and deep learning for brain stroke detection-A review of recent advancements and future prospects. Computer Methods Programs in Biomedicine, 197, 105728.
19. Keskin, R., Çaylı, Ö., Moral, Ö. T., Kılıc, V., & Onan, A. (2021). A benchmark for feature-injection architectures in image captioning. Avrupa Bilim ve Teknoloji Dergisi, 31, 461-468.
20. Keskin, R., Moral, Ö. T., Kılıç, V., & Onan, A. (2021). Multi-gru based automated image captioning for smartphones. Paper presented at the 2021 29th Signal Processing and Communications Applications Conference (SIU).
21. Khezrpour, S., Seyedarabi, H., Razavi, S. N., & Farhoudi, M. (2022). Automatic segmentation of the brain stroke lesions from MR flair scans using improved U-net framework. Biomedical Signal Processing Control, 78, 103978.
22. Kilic, B., Dogan, V., Kilic, V., & Kahyaoglu, L. N. (2022). Colorimetric food spoilage monitoring with carbon dot and UV light reinforced fish gelatin films using a smartphone application. International Journal of Biological Macromolecules, 209, 1562-1572.
23. Kilic, V., & Şen, M. (2019). Smartphone-based Colorimetric Analysis for the Detection of H 2 O 2 Using a ¼PAD. Paper presented at the 2019 Medical Technologies Congress (TIPTEKNO).
24. Kılıc, V. J. (2021). Deep gated recurrent unit for smartphone-based image captioning. Sakarya University Journal of Computer Information Sciences, 4(2), 181-191.
25. Kılıç, V. Yapay Zeka Tabanlı Akıllı Telefon Uygulaması ile Kan Şekeri Tahmini. Avrupa Bilim ve Teknoloji Dergisi, 26, 289-294.
26. Kılıç, V., Mercan, Ö. B., Tetik, M., Kap, Ö., & Horzum, N. (2022). Non-enzymatic colorimetric glucose detection based on Au/Ag nanoparticles using smartphone and machine learning. Analytical Sciences, 38(2), 347-358.
27. Kirshner, H., & Schrag, M. (2021). Management of intracerebral hemorrhage: update and future therapies. Current Neurology Neuroscience Reports, 21, 1-5.
28. Koç, U., Sezer, E. A., Özkaya, Y. A., Yarbay, Y., Taydaş, O., Ayyıldız, V. A., . . . Beşler, M. S. (2022). Artificial Intelligence in Healthcare Competition (Teknofest-2021): Stroke Data Set. The Eurasian journal of medicine, 54(3), 248.
29. Kökten, A., & Kılıç, V. (2021). Detection of COVID-19 Cases with Fuzzy Classifiers Using Chest Computed Tomography. Avrupa Bilim ve Teknoloji Dergisi, 26, 68-72.
30. Kumar, A., et al. (2020). CSNet: A new DeepNet framework for ischemic stroke lesion segmentation. Computer Methods and Programs in Biomedicine, 193.
31. Liu, L., Kurgan, L., Wu, F.-X., & Wang, J. J. (2020). Attention convolutional neural network for accurate segmentation and quantification of lesions in ischemic stroke disease. Medical Image Analysis, 65, 101791.
32. Mercan, Ö. B., Doğan, V., & Kılıç, V. (2020). Time Series Analysis based Machine Learning Classification for Blood Sugar Levels. Paper presented at the 2020 Medical Technologies Congress (TIPTEKNO).
33. Mercan, Ö. B., & Kılıç, V. (2020). Deep Learning based Colorimetric Classification of Glucose with Au-Ag nanoparticles using Smartphone. Paper presented at the 2020 Medical Technologies Congress (TIPTEKNO).
34. Mercan, Ö. B., & Kılıç, V. (2021). Fuzzy classifier based colorimetric quantification using a smartphone. Paper presented at the Intelligent and Fuzzy Techniques: Smart and Innovative Solutions: Proceedings of the INFUS 2020 Conference, Istanbul, Turkey, July 21-23, 2020.
35. Oktay, O., Schlemper, J., Folgoc, L. L., Lee, M., Heinrich, M., Misawa, K., . . . Kainz, B. J. (2018). Attention u-net: Learning where to look for the pancreas. arXiv preprint arXiv.
36. Palaz, Z., Doğan, V., & Kılıç, V. J. (2021). Smartphone-based Multi-parametric Glucose Prediction using Recurrent Neural Networks. Avrupa Bilim ve Teknoloji Dergisi, 32, 1168-1174.
37. Rajinikanth, V., Fernandes, S. L., Bhushan, B., & Sunder, N. R. (2018). Segmentation and analysis of brain tumor using Tsallis entropy and regularised level set. In Proceedings of 2nd International Conference on Micro-Electronics, Electromagnetics and Telecommunications: ICMEET 2016, (pp. 313-321).
38. Ronneberger, O., Fischer, P., & Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. Paper presented at the Medical Image Computing and Computer-Assisted Intervention–MICCAI 2015: 18th International Conference, Munich, Germany, October 5-9, 2015, Proceedings, Part III 18.
39. Sayraci, B., Agrali, M., & Kilic, V. J. (2023). Artificial Intelligence Based Instance-Aware Semantic Lobe Segmentation on Chest Computed Tomography Images. Avrupa Bilim ve Teknoloji Dergisi, 46, 109-115.
40. Simonyan, K., & Zisserman, A. J. (2014). Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:.
41. Şen, M., Yüzer, E., Doğan, V., Avcı, İ., Ensarioğlu, K., Aykaç, A., . . . Kılıç, V. J. (2022). Colorimetric detection of H2O2 with Fe3O4@ Chi nanozyme modified µPADs using artificial intelligence. Microchimica Acta, 189(10), 373.
42. Tursynova, A., Omarov, B., Sakhipov, A., & Tukenova, N. J. (2022). Brain Stroke Lesion Segmentation Using Computed Tomography Images based on Modified U-Net Model with ResNet Blocks. International Journal of Online Biomedical Engineering, 18(13).
43. Wu, J., & Tang, X. (2019). Brain segmentation based on multi-atlas guided 3D fully convolutional network ensembles. arXiv preprint arXiv, 1901.01381.
44. Yüzer, E., Doğan, V., Kılıç, V., & Şen, M. J. (2022). Smartphone embedded deep learning approach for highly accurate and automated colorimetric lactate analysis in sweat. Sensors Actuators B: Chemical, 371, 132489.
45. Zhang, Z., Liu, Q., & Wang, Y. J. (2018). Road extraction by deep residual u-net. IEEE Geoscience Remote Sensing Letters, 15(5), 749-753.