Research Article
BibTex RIS Cite

Mixup Veri Artırma Yöntemi ile Retina Damar Bölütlemesi

Year 2022, Volume: 5 Issue: 1, 41 - 50, 28.04.2022
https://doi.org/10.54537/tusebdergisi.1083833

Abstract

Derin sinir ağı modellerinin aşırı uyum sorununun azaltılması için veri artırma yöntemlerine başvurulmaktadır. 2018 yılında bir veri artırma yöntemi olan mixup tanıtılmıştır ve devam eden yıllarda farklı organ ve görüntü modalitelerinde yapılan çalışmalar ile mixup yönteminin model bölütleme yeteneği üzerindeki etkisi incelenmiştir. Mixup yönteminin tarayıcı lazer oftalmoskop ile elde edilmiş fundus görüntülerinde retina damar bölütlemesi için kullanımına yönelik bir çalışmaya rastlanmamıştır. Bu çalışmanın amacı, IOSTAR veri kümesi görüntülerinde U-Net modeli ile gerçekleştirilen retina damar bölütlemesine mixup yönteminin etkisinin incelenmesidir. Bu doğrultuda yatay yansıtma, görüntünün rastgele bir alanını kırpma, çevirme gibi veri artırma işlemlerinin görüntülere uygulandığı bir geleneksel grup; geleneksel yöntem ile oluşturulmuş görüntülere ek olarak lambda 0,2 veya 0,5 değerlerine göre mixup yönteminin uygulandığı iki farklı grup; lambda 0,2 veya 0,5 değerlerine göre sadece mixup yönteminin uygulandığı iki farklı grup olmak üzere beş farklı veri grubu oluşturulmuştur. Doğruluk, duyarlılık, özgüllük, Dice ve Jaccard ölçütlerine göre değerlendirmeler yapılmıştır. Geleneksel veri artırma yöntemleriyle karşılaştırıldığında, U-Net modelinin retina damar bölütleme yeteneğine mixup veri artırma yönteminin iyileşme sağlamadığı görülmüştür.

References

  • Arpacı, S. A., & Varlı, S. (2020, Ekim). Farklılaştırılmış U-Net ağı ile retina damar bölütlemesi. Sözel bildiri, 2020 28. IEEE Sinyal İşleme ve İletişim Uygulamaları Kurultayı (SIU), Gaziantep, Türkiye.
  • Arpacı, S. A., & Varlı, S. (2021a, Haziran). EncU-Net: Dermoskopik görüntü bölütlemesi için modifiye edilmiş U-Net. Sözel bildiri, 2021 29. IEEE Sinyal İşleme ve İletişim Uygulamaları Kurultayı (SIU), İstanbul, Türkiye.
  • Arpacı, S. A. & Varlı, S. (2021b). LUPU-Net: a new improvement proposal for encoder-decoder architecture. International Advanced Researches and Engineering Journal, 5 (3), 352-361. doi: 10.35860/iarej.939243
  • Brea, L.M., Jesus, D.A., Klein, S., & Walsum, T.V. (2020, July). Deep learning-based retinal vessel segmentation with cross-modal evaluation. Poster, Third Conference on Medical Imaging with Deep Learning (MIDL), Montreal, QC, Canada.
  • Cai, Y., Li, Y., Gao, X., & Guo, Y. (2019, October). Retinal vessel segmentation method based on improved deep U-Net. Sözel bildiri, Chinese Conference on Biometric Recognition, Zhuzhou, China.
  • Cheng, Y., Ma, M., Zhang, L., Jin, C., Ma, L., & Zhou, Y. (2020). Retinal blood vessel segmentation based on densely connected U-net. Mathematical Biosciences and Engineering, 17 (4), 3088-3108. doi:10.3934/mbe.2020175
  • Eaton-Rosen, Z., Bragman, F.J., Ourselin, S., & Cardoso, M.J. (2018, July). Improving data augmentation for medical image segmentation. Sözel bildiri, International Conference on Medical Imaging with Deep Learning, Amsterdam, Netherlands.
  • Gazda, M., Bugata, P., Gazda, J., Hubacek, D., Hresko, D. J., & Drotar, P. (2021). Mixup augmentation for kidney and kidney tumor segmentation. https://openreview.net/forum?id=GvUsPesMtmC adresinden elde edildi.
  • Guo, C., Szemenyei, M., Yi, Y., Xue, Y., Zhou, W., & Li, Y. (2020, May). Dense residual network for retinal vessel segmentation. Poster, 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Barcelona, Spain.
  • He, K., Zhang, X., Ren, S., & Sun, J. (2016, June). Deep residual learning for image recognition. Sözel bildiri, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, USA.
  • Huang, G., Liu, Z., Van Der Maaten, L., & Weinberger, K. Q. (2017, July). Densely connected convolutional networks. Sözel bildiri, IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, Hawaii.
  • Hu, J., Shen, L., & Sun, G. (2018, June). Squeeze-and-excitation networks. Sözel bildiri, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Salt Lake City, USA.
  • Isaksson, L. J., Summers, P., Raimondi, S., Gandini, S., Bhalerao, A., Marvaso, G., et. al. (2021). Mixup (sample pairing) can improve the performance of deep segmentation networks. Journal of Artificial Intelligence and Soft Computing Research, 12 (1), 29-39. doi:10.2478/jaiscr-2022-0003
  • Khan, K. B., Khaliq, A. A., Jalil, A., Iftikhar, M. A., Ullah, N., Aziz, M., ve diğerleri. (2018). A review of retinal blood vessels extraction techniques: Challenges, taxonomy, and future trends. Pattern Analysis and Applications, 22 (3), 767-802. doi:10.1007/s10044-018-0754-8
  • Kingma, D., & Ba, J. (2015). Adam: A method for stochastic optimization. https://arxiv.org/abs/1412.6980v5 adresinden elde edildi.
  • Li, Q., Fan, S., & Chen, C. (2019). An intelligent segmentation and diagnosis method for diabetic retinopathy based on improved U-NET network. Journal of Medical Systems, 43 (9), 304. doi:10.1007/s10916-019-1432-0
  • Li, X., Jiang, Y., Li, M., & Yin, S. (2021). Lightweight attention convolutional neural network for retinal vessel image segmentation. IEEE Transactions on Industrial Informatics, 17 (3), 1958-1967. doi:10.1109/tii.2020.2993842
  • Meyer, M. I., Costa, P., Galdran, A., Mendonça, A. M., & Campilho, A. (2017). A deep neural network for vessel segmentation of scanning laser ophthalmoscopy images. https://link.springer.com/content/pdf/10.1007%2F978-3-319-59876-5_56.pdf adresinden elde edildi.
  • Nishio, M., Noguchi, S., & Fujimoto, K. (2020). Automatic pancreas segmentation using coarse-scaled 2D model of deep learning: Usefulness of data augmentation and deep U-Net. Applied Sciences, 10 (10), 3360. doi:10.3390/app10103360
  • Noguchi, S., Nishio, M., Yakami, M., Nakagomi, K., & Togashi, K. (2020). Bone segmentation on whole-body CT using convolutional neural network with novel data augmentation techniques. Computers in Biology and Medicine, 121, 103767. doi:10.1016/j.compbiomed.2020.103767
  • Pachade, S., Porwal, P., Kokare, M., Giancardo, L., & Meriaudeau, F. (2020). Retinal vasculature segmentation and measurement framework for color fundus and SLO Images. Biocybernetics and Biomedical Engineering, 40 (3), 865-900. doi:10.1016/j.bbe.2020.03.001
  • Ronneberger, O., Fischer, P., & Brox, T. (2015). U-Net: Convolutional networks for biomedical image segmentation. https://arxiv.org/abs/1505.04597 adresinden elde edildi.
  • Sureshjani, S., Ockeloen, I., Zhang, J., & Ter Haar Romeny, B. (2015, July). Biologically-inspired supervised vasculature segmentation in SLO retinal fundus images. Sözel bildiri, 12th International Conference, ICIAR 2015, Niagara Falls, Canada.
  • Terasaki, H., Sonoda, S., Tomita, M., & Sakamoto, T. (2021). Recent advances and clinical application of color scanning laser ophthalmoscope. Journal of Clinical Medicine, 10 (4), 718. doi:10.3390/jcm10040718
  • Xiao, X., Lian, S., Luo, Z., & Li, S. (2018, October). Weighted Res-UNet for high-quality retina vessel segmentation. Sözel bildiri, 2018 9th International Conference on Information Technology in Medicine and Education (ITME), Hangzhou, China.
  • Zhang, J., Dashtbozorg, B., Bekkers, E., Pluim, J. P., Duits, R., & Ter Haar Romeny, B. M. (2016). Robust retinal vessel segmentation via locally adaptive derivative frames in orientation scores. IEEE Transactions on Medical Imaging, 35 (12), 2631-2644. doi:10.1109/tmi.2016.2587062
  • Zhang, H., Cissé, M., Dauphin, Y., & Lopez-Paz, D. (2018, April). mixup: Beyond empirical risk minimization. Poster, 6th International Conference on Learning Representations (ICLR), Vancouver, BC, Canada.
Year 2022, Volume: 5 Issue: 1, 41 - 50, 28.04.2022
https://doi.org/10.54537/tusebdergisi.1083833

Abstract

References

  • Arpacı, S. A., & Varlı, S. (2020, Ekim). Farklılaştırılmış U-Net ağı ile retina damar bölütlemesi. Sözel bildiri, 2020 28. IEEE Sinyal İşleme ve İletişim Uygulamaları Kurultayı (SIU), Gaziantep, Türkiye.
  • Arpacı, S. A., & Varlı, S. (2021a, Haziran). EncU-Net: Dermoskopik görüntü bölütlemesi için modifiye edilmiş U-Net. Sözel bildiri, 2021 29. IEEE Sinyal İşleme ve İletişim Uygulamaları Kurultayı (SIU), İstanbul, Türkiye.
  • Arpacı, S. A. & Varlı, S. (2021b). LUPU-Net: a new improvement proposal for encoder-decoder architecture. International Advanced Researches and Engineering Journal, 5 (3), 352-361. doi: 10.35860/iarej.939243
  • Brea, L.M., Jesus, D.A., Klein, S., & Walsum, T.V. (2020, July). Deep learning-based retinal vessel segmentation with cross-modal evaluation. Poster, Third Conference on Medical Imaging with Deep Learning (MIDL), Montreal, QC, Canada.
  • Cai, Y., Li, Y., Gao, X., & Guo, Y. (2019, October). Retinal vessel segmentation method based on improved deep U-Net. Sözel bildiri, Chinese Conference on Biometric Recognition, Zhuzhou, China.
  • Cheng, Y., Ma, M., Zhang, L., Jin, C., Ma, L., & Zhou, Y. (2020). Retinal blood vessel segmentation based on densely connected U-net. Mathematical Biosciences and Engineering, 17 (4), 3088-3108. doi:10.3934/mbe.2020175
  • Eaton-Rosen, Z., Bragman, F.J., Ourselin, S., & Cardoso, M.J. (2018, July). Improving data augmentation for medical image segmentation. Sözel bildiri, International Conference on Medical Imaging with Deep Learning, Amsterdam, Netherlands.
  • Gazda, M., Bugata, P., Gazda, J., Hubacek, D., Hresko, D. J., & Drotar, P. (2021). Mixup augmentation for kidney and kidney tumor segmentation. https://openreview.net/forum?id=GvUsPesMtmC adresinden elde edildi.
  • Guo, C., Szemenyei, M., Yi, Y., Xue, Y., Zhou, W., & Li, Y. (2020, May). Dense residual network for retinal vessel segmentation. Poster, 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Barcelona, Spain.
  • He, K., Zhang, X., Ren, S., & Sun, J. (2016, June). Deep residual learning for image recognition. Sözel bildiri, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, USA.
  • Huang, G., Liu, Z., Van Der Maaten, L., & Weinberger, K. Q. (2017, July). Densely connected convolutional networks. Sözel bildiri, IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, Hawaii.
  • Hu, J., Shen, L., & Sun, G. (2018, June). Squeeze-and-excitation networks. Sözel bildiri, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Salt Lake City, USA.
  • Isaksson, L. J., Summers, P., Raimondi, S., Gandini, S., Bhalerao, A., Marvaso, G., et. al. (2021). Mixup (sample pairing) can improve the performance of deep segmentation networks. Journal of Artificial Intelligence and Soft Computing Research, 12 (1), 29-39. doi:10.2478/jaiscr-2022-0003
  • Khan, K. B., Khaliq, A. A., Jalil, A., Iftikhar, M. A., Ullah, N., Aziz, M., ve diğerleri. (2018). A review of retinal blood vessels extraction techniques: Challenges, taxonomy, and future trends. Pattern Analysis and Applications, 22 (3), 767-802. doi:10.1007/s10044-018-0754-8
  • Kingma, D., & Ba, J. (2015). Adam: A method for stochastic optimization. https://arxiv.org/abs/1412.6980v5 adresinden elde edildi.
  • Li, Q., Fan, S., & Chen, C. (2019). An intelligent segmentation and diagnosis method for diabetic retinopathy based on improved U-NET network. Journal of Medical Systems, 43 (9), 304. doi:10.1007/s10916-019-1432-0
  • Li, X., Jiang, Y., Li, M., & Yin, S. (2021). Lightweight attention convolutional neural network for retinal vessel image segmentation. IEEE Transactions on Industrial Informatics, 17 (3), 1958-1967. doi:10.1109/tii.2020.2993842
  • Meyer, M. I., Costa, P., Galdran, A., Mendonça, A. M., & Campilho, A. (2017). A deep neural network for vessel segmentation of scanning laser ophthalmoscopy images. https://link.springer.com/content/pdf/10.1007%2F978-3-319-59876-5_56.pdf adresinden elde edildi.
  • Nishio, M., Noguchi, S., & Fujimoto, K. (2020). Automatic pancreas segmentation using coarse-scaled 2D model of deep learning: Usefulness of data augmentation and deep U-Net. Applied Sciences, 10 (10), 3360. doi:10.3390/app10103360
  • Noguchi, S., Nishio, M., Yakami, M., Nakagomi, K., & Togashi, K. (2020). Bone segmentation on whole-body CT using convolutional neural network with novel data augmentation techniques. Computers in Biology and Medicine, 121, 103767. doi:10.1016/j.compbiomed.2020.103767
  • Pachade, S., Porwal, P., Kokare, M., Giancardo, L., & Meriaudeau, F. (2020). Retinal vasculature segmentation and measurement framework for color fundus and SLO Images. Biocybernetics and Biomedical Engineering, 40 (3), 865-900. doi:10.1016/j.bbe.2020.03.001
  • Ronneberger, O., Fischer, P., & Brox, T. (2015). U-Net: Convolutional networks for biomedical image segmentation. https://arxiv.org/abs/1505.04597 adresinden elde edildi.
  • Sureshjani, S., Ockeloen, I., Zhang, J., & Ter Haar Romeny, B. (2015, July). Biologically-inspired supervised vasculature segmentation in SLO retinal fundus images. Sözel bildiri, 12th International Conference, ICIAR 2015, Niagara Falls, Canada.
  • Terasaki, H., Sonoda, S., Tomita, M., & Sakamoto, T. (2021). Recent advances and clinical application of color scanning laser ophthalmoscope. Journal of Clinical Medicine, 10 (4), 718. doi:10.3390/jcm10040718
  • Xiao, X., Lian, S., Luo, Z., & Li, S. (2018, October). Weighted Res-UNet for high-quality retina vessel segmentation. Sözel bildiri, 2018 9th International Conference on Information Technology in Medicine and Education (ITME), Hangzhou, China.
  • Zhang, J., Dashtbozorg, B., Bekkers, E., Pluim, J. P., Duits, R., & Ter Haar Romeny, B. M. (2016). Robust retinal vessel segmentation via locally adaptive derivative frames in orientation scores. IEEE Transactions on Medical Imaging, 35 (12), 2631-2644. doi:10.1109/tmi.2016.2587062
  • Zhang, H., Cissé, M., Dauphin, Y., & Lopez-Paz, D. (2018, April). mixup: Beyond empirical risk minimization. Poster, 6th International Conference on Learning Representations (ICLR), Vancouver, BC, Canada.
There are 27 citations in total.

Details

Primary Language Turkish
Subjects Computer Software
Journal Section Araştırma Makaleleri
Authors

Saadet Aytaç Arpacı

Songül Varlı 0000-0002-1786-6869

Publication Date April 28, 2022
Published in Issue Year 2022 Volume: 5 Issue: 1

Cite

APA Arpacı, S. A., & Varlı, S. (2022). Mixup Veri Artırma Yöntemi ile Retina Damar Bölütlemesi. Türkiye Sağlık Enstitüleri Başkanlığı Dergisi, 5(1), 41-50. https://doi.org/10.54537/tusebdergisi.1083833