Ahmet Furkan BAYRAM, Alperen DERİN, Caglar GURKAN, Abdulkadir BUDAK, Hakan KARATAŞ

İskemik İnme Lezyon Segmentasyonunda Segmentasyon Ağlarının ve Kayıp Fonksiyonlarının Etkilerinin Analizi

2020'de serebrovasküler hastalıklardan her altı ölümden birinin nedeni inmeydi. Amerika Birleşik Devletleri'nde (ABD) her 40 saniyede bir inme vakası görülmektedir. Her 3.5 dakikada bir insan inmeden hayatını kaybetmektedir. ABD'de yılda toplamda 795.000'den fazla inme vakası meydana gelmektedir. Bu çalışma, beyinde oluşan iskemik inme lezyonunu tespit etmeyi amaçlamaktadır. 43 hastanın 82 Manyetik Rezonans görüntüsünü içeren İskemik İnme Lezyon Segmentasyonu (ISLES) 2017 veri seti kullanıldı. UNet, Attention UNet, Residual UNet, Attention Residual UNet ve Residual UNet++ segmentasyon ağları test edilmiştir. Ayrıca Cross Entropy, Dice, IoU, Tversky, Focal Tversky ve bunların bileşik formları incelenmiştir. Attention UNet üzerinde test edilen IoU kayıp fonksiyonu 0.766 Zar skoru, 0.621 IoU skoru, 0.730 duyarlılık, 0.997 özgüllük, 0.805 kesinlik ve 0.993 doğruluk ile en iyi performansı elde etmiştir.

Anahtar Kelimeler:

İnme, Segmentasyon, Yapay zeka, Makine öğrenmesi, Derin öğrenme

Analysis of the Effects of Segmentation Networks and Loss Functions in Ischemic Stroke Lesion Segmentation

Stroke was the cause of one out of every six deaths from cerebrovascular disease in 2020. A stroke occurs in the United States (US) every 40 seconds. Every 3.5 minutes, people die of a stroke. More than total 795,000 stroke cases occur yearly in the US. This study aims to detect the ischemic stroke lesion that occurs in the brain. The Ischemic Stroke Lesion Segmentation (ISLES) 2017 data set, which includes 82 Magnetic Resonance images of 43 patients, was used. The UNet, Attention UNet, Residual UNet, Attention Residual UNet, and Residual UNet++ segmentation networks were tested. Moreover, Cross Entropy, Dice, IoU, Tversky, Focal Tversky, and their compound forms were analyzed. The IoU loss function tested on Attention UNet achieved the best performance with the dice score of 0.766, the IoU score of 0.621, the sensitivity of 0.730, the specificity of 0.997, the precision of 0.805, and the accuracy of 0.993.

Keywords:

Stroke, Segmentation, Artificial intelligence, Machine learning, Deep learning,

PDF

___

Centers for Disease Control and Prevention. (2020). Underlying Cause of Death, 1999-2020 Request. CDC WONDER Online Database. https://wonder.cdc.gov/ucd-icd10.html
Tsao, C. W., Aday, A. W., Almarzooq, Z. I., Alonso, A., Beaton, A. Z., Bittencourt, M. S., Boehme, A. K., Buxton, A. E., Carson, A. P., Commodore-Mensah, Y., Elkind, M. S. V., Evenson, K. R., Eze-Nliam, C., Ferguson, J. F., Generoso, G., Ho, J. E., Kalani, R., Khan, S. S., Kissela, B. M., … Martin, S. S. (2022). Heart Disease and Stroke Statistics-2022 Update: A Report From the American Heart Association. Circulation, 145(8), e153–e639. https://doi.org/10.1161/CIR.0000000000001052
Kadry, S., Damasevicius, R., Taniar, D., Rajinikanth, V., & Lawal, I. A. (2021, March 25). U-Net Supported Segmentation of Ischemic-Stroke-Lesion from Brain MRI Slices. Proceedings of 2021 IEEE 7th International Conference on Bio Signals, Images and Instrumentation, ICBSII 2021. https://doi.org/10.1109/ICBSII51839.2021.9445126
Shin, H., Agyeman, R., Rafiq, M., Chang, M. C., & Choi, G. S. (2022). Automated segmentation of chronic stroke lesion using efficient U-Net architecture. Biocybernetics and Biomedical Engineering, 42(1), 285–294. https://doi.org/10.1016/j.bbe.2022.01.002
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. SSRN Electronic Journal, 78, 103978. https://doi.org/10.2139/ssrn.4015024
Soltanpour, M., Greiner, R., Boulanger, P., & Buck, B. (2021). Improvement of automatic ischemic stroke lesion segmentation in CT perfusion maps using a learned deep neural network. Computers in Biology and Medicine, 137, 104849. https://doi.org/10.1016/j.compbiomed.2021.104849
Ou, Y., Yuan, Y., Huang, X., Wong, K., Volpi, J., Wang, J. Z., & Wong, S. T. C. (2021). LambdaUNet: 2.5D Stroke Lesion Segmentation of Diffusion-Weighted MR Images. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 12901 LNCS, 731–741. https://doi.org/10.1007/978-3-030-87193-2_69
HarisIqbal88/PlotNeuralNet: Latex code for making neural networks diagrams. (n.d.). Retrieved September 7, 2022, from https://github.com/HarisIqbal88/PlotNeuralNet