PERBANDINGAN ARSITEKTUR CONVOLUTIONAL NEURAL NETWORK UNTUK KLASIFIKASI FUNDUS

Wahyudi Setiawan

Abstract


Pada artikel ini membahas tentang perbandingan arsitektur Convolutional Neural Network (CNN) untuk klasifikasi citra fundus. Arsitektur CNN yang diujicobakan yaitu AlexNet, Visual Geometry Group (VGG) 16, VGG19, Residual Network (ResNet) 50, ResNet101, GoogleNet, Inception-V3, InceptionResNetV2 dan Squeezenet. Citra ujicoba menggunakan fundus retina utnuk mengklasifikasi 2 kelas yaitu normal dan neovaskularisasi. Citra dilakukan preprosesing yaitu dengan membaginya menjadi 16 bagian yang sama. Skenario ujicoba menggunakan 2 tahap yaitu, pertama, menggunakan CNN tanpa optimasi tambahan, kedua, CNN menggunakan optimasi Gradient Descent. Hasil ujicoba pada kedua skenario menunjukkan arsitektur terbaik yaitu VGG19 dan VGG16. Ujicoba tahap pertama menghasilkan sensitivitas, spesifisitas dan akurasi yaitu 87,8%, 90,7% dan 89,3%. Untuk ujicoba tahap kedua sensitivitas, spesifisitas dan akurasi yaitu 94,2%, 90,4% dan 92,31%.


References


Lechun, “gradient based learning applied to document recognition,” proc.of IEEE, no. november, pp. 1–46, 1998.

A. Krizhevsky, I. Sutskever, and G. E. Hinton, “ImageNet Classification with Deep Convolutional Neural Networks,” 2012.

K. Simonyan and A. Zisserman, “Very Deep Convolutional Networks for Large-Scale Image Recognition,” in ICLR, 2015, pp. 1–14.

C. Szegedy, V. Vanhouke, S. Ioffe, and J. Shlens, “Rethinking the Inception Architecture for Computer Vision Christian,” IEEE Explor., pp. 2818–2826, 2015.

K. He and J. Sun, “Deep Residual Learning for Image Recognition,” IEEE Xplore, pp. 1–9, 2015.

C. Szegedy, S. Ioffe, and V. Vanhoucke, “Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning,” pp. 1–12, 2016.

F. N. Iandola, S. Han, M. W. Moskewicz, K. Ashraf, W. J. Dally, and K. Keutzer, “SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size,” in ICLR, 2017, pp. 1–13.

W. Setiawan, M. . Utoyo, and R. Rulaningtyas, “Classification of neovascularization using convolutional neural network model,” TELKOMNIKA, vol. 17, no. 1, pp. 463–473, 2019.

S. Yu, D. Xiao, and Y. Kanagasingam, “Exudate Detection for Diabetic Retinopathy With Convolutional Neural Networks,” pp. 1744–1747, 2017.

A. Kwasigroch, B. Jarzembinski, and M. Grochowski, “Deep CNN based decision support system for detection and assessing the stage of diabetic retinopathy,” 2018 Int. Interdiscip. PhD Work., pp. 111–116, 2018.

R. Ghosh, “Automatic Detection and Classification of Diabetic Retinopathy stages using CNN,” 2017.

E. Decencière et al., “Feedback On A Publicly Distributed Image Database : The Messidor Database,” Image Anal Stereol, vol. 33, pp. 231–234, 2014.

S. Ruder, “An overview of gradient descent optimization,” 2017.

G. E. Hinton, N. Srivastava, and K. Swersky, “Lecture 6a- overview of mini-batch gradient descent,” COURSERA Neural Networks Mach. Learn., p. 31, 2012.

D. P. Kingma and J. Ba, “Adam: A Method for Stochastic Optimization,” in ICLR, 2015, pp. 1–15.




DOI: https://doi.org/10.21107/simantec.v7i2.6551

Refbacks

  • There are currently no refbacks.


Copyright (c) 2020 Wahyudi Setiawan

Indexed By