TY - GEN
T1 - Multiscale network followed network model for retinal vessel segmentation
AU - Wu, Yicheng
AU - Xia, Yong
AU - Song, Yang
AU - Zhang, Yanning
AU - Cai, Weidong
N1 - Publisher Copyright:
© Springer Nature Switzerland AG 2018.
PY - 2018
Y1 - 2018
N2 - The shape of retinal blood vessels plays a critical role in the early diagnosis of diabetic retinopathy. However, it remains challenging to segment accurately the blood vessels, particularly the capillaries, in color retinal images. In this paper, we propose the multiscale network followed network (MS-NFN) model to address this issue. This model consists of an ‘up-pool’ NFN submodel and a ‘pool-up’ NFN submodel, in which max-pooling layers and up-sampling layers can generate multiscale feature maps. In each NFN, the first multiscale network converts an image patch into a probabilistic retinal vessel map, and the following multiscale network further refines the map. The refined probabilistic retinal vessel maps produced by both NFNs are averaged to construct the segmentation result. We evaluated this model on the digital retinal images for vessel extraction (DRIVE) dataset and the child heart and health study dataset. Our results indicate that the NFN structure we designed is able to produce performance gain and the proposed MS-NFN model achieved the state-of-the-art retinal vessel segmentation accuracy on both datasets.
AB - The shape of retinal blood vessels plays a critical role in the early diagnosis of diabetic retinopathy. However, it remains challenging to segment accurately the blood vessels, particularly the capillaries, in color retinal images. In this paper, we propose the multiscale network followed network (MS-NFN) model to address this issue. This model consists of an ‘up-pool’ NFN submodel and a ‘pool-up’ NFN submodel, in which max-pooling layers and up-sampling layers can generate multiscale feature maps. In each NFN, the first multiscale network converts an image patch into a probabilistic retinal vessel map, and the following multiscale network further refines the map. The refined probabilistic retinal vessel maps produced by both NFNs are averaged to construct the segmentation result. We evaluated this model on the digital retinal images for vessel extraction (DRIVE) dataset and the child heart and health study dataset. Our results indicate that the NFN structure we designed is able to produce performance gain and the proposed MS-NFN model achieved the state-of-the-art retinal vessel segmentation accuracy on both datasets.
KW - Fully convolutional network
KW - Network followed network
KW - Retinal vessel segmentation
UR - http://www.scopus.com/inward/record.url?scp=85054074994&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-00934-2_14
DO - 10.1007/978-3-030-00934-2_14
M3 - 会议稿件
AN - SCOPUS:85054074994
SN - 9783030009335
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 119
EP - 126
BT - Medical Image Computing and Computer Assisted Intervention – MICCAI 2018 - 21st International Conference, 2018, Proceedings
A2 - Fichtinger, Gabor
A2 - Davatzikos, Christos
A2 - Alberola-López, Carlos
A2 - Frangi, Alejandro F.
A2 - Schnabel, Julia A.
PB - Springer Verlag
T2 - 21st International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2018
Y2 - 16 September 2018 through 20 September 2018
ER -
