TY - GEN
T1 - Automatic feature learning for glaucoma detection based on deep learning
AU - Chen, Xiangyu
AU - Xu, Yanwu
AU - Yan, Shuicheng
AU - Wong, Damon Wing Kee
AU - Wong, Tien Yin
AU - Liu, Jiang
N1 - Publisher Copyright:
© Springer International Publishing Switzerland 2015.
PY - 2015
Y1 - 2015
N2 - Glaucoma is a chronic and irreversible eye disease in which the optic nerve is progressively damaged, leading to deterioration in vision and quality of life. In this paper, we present an Automatic feature Learning for glAucomaDetection based onDeep LearnINg (ALADDIN),with deep convolutional neural network (CNN) for feature learning. Different from the traditional convolutional layer that uses linear filters followed by a nonlinear activation function to scan the input, the adopted network embeds micro neural networks (multilayer perceptron) with more complex structures to abstract the data within the receptive field. Moreover, a contextualizing deep learning structure is proposed in order to obtain a hierarchical representation of fundus images to discriminate between glaucoma and non-glaucoma pattern,where the network takes the outputs fromother CNN as the context information to boost the performance. Extensive experiments are performed on the ORIGA and SCES datasets. The results showarea under curve (AUC) of the receiver operating characteristic curve in glaucoma detection at 0.838 and 0.898 in the two databases,much better than state-of-the-art algorithms. The method could be used for glaucoma diagnosis.
AB - Glaucoma is a chronic and irreversible eye disease in which the optic nerve is progressively damaged, leading to deterioration in vision and quality of life. In this paper, we present an Automatic feature Learning for glAucomaDetection based onDeep LearnINg (ALADDIN),with deep convolutional neural network (CNN) for feature learning. Different from the traditional convolutional layer that uses linear filters followed by a nonlinear activation function to scan the input, the adopted network embeds micro neural networks (multilayer perceptron) with more complex structures to abstract the data within the receptive field. Moreover, a contextualizing deep learning structure is proposed in order to obtain a hierarchical representation of fundus images to discriminate between glaucoma and non-glaucoma pattern,where the network takes the outputs fromother CNN as the context information to boost the performance. Extensive experiments are performed on the ORIGA and SCES datasets. The results showarea under curve (AUC) of the receiver operating characteristic curve in glaucoma detection at 0.838 and 0.898 in the two databases,much better than state-of-the-art algorithms. The method could be used for glaucoma diagnosis.
UR - http://www.scopus.com/inward/record.url?scp=84951789546&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-24574-4_80
DO - 10.1007/978-3-319-24574-4_80
M3 - Conference contribution
AN - SCOPUS:84951789546
SN - 9783319245737
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 669
EP - 677
BT - Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015 - 18th International Conference, Proceedings
A2 - Frangi, Alejandro F.
A2 - Navab, Nassir
A2 - Hornegger, Joachim
A2 - Wells, William M.
PB - Springer Verlag
T2 - 18th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2015
Y2 - 5 October 2015 through 9 October 2015
ER -