TY - GEN
T1 - Recurrent Neural Networks for Aortic Image Sequence Segmentation with Sparse Annotations
AU - Bai, Wenjia
AU - Suzuki, Hideaki
AU - Qin, Chen
AU - Tarroni, Giacomo
AU - Oktay, Ozan
AU - Matthews, Paul M.
AU - Rueckert, Daniel
N1 - Funding Information:
Acknowledgements. This research has been conducted using the UK Biobank Resource under Application Number 18545. This work is supported by the SmartHeart EPSRC Programme Grant (EP/P001009/1). We would like to acknowledge NVIDIA Corporation for donating a Titan Xp for this research. P.M.M. thanks the Edmond J. Safra Foundation, Lily Safra and the UK Dementia Research Institute for their generous support.
Publisher Copyright:
© 2018, Springer Nature Switzerland AG.
PY - 2018
Y1 - 2018
N2 - Segmentation of image sequences is an important task in medical image analysis, which enables clinicians to assess the anatomy and function of moving organs. However, direct application of a segmentation algorithm to each time frame of a sequence may ignore the temporal continuity inherent in the sequence. In this work, we propose an image sequence segmentation algorithm by combining a fully convolutional network with a recurrent neural network, which incorporates both spatial and temporal information into the segmentation task. A key challenge in training this network is that the available manual annotations are temporally sparse, which forbids end-to-end training. We address this challenge by performing non-rigid label propagation on the annotations and introducing an exponentially weighted loss function for training. Experiments on aortic MR image sequences demonstrate that the proposed method significantly improves both accuracy and temporal smoothness of segmentation, compared to a baseline method that utilises spatial information only. It achieves an average Dice metric of 0.960 for the ascending aorta and 0.953 for the descending aorta.
AB - Segmentation of image sequences is an important task in medical image analysis, which enables clinicians to assess the anatomy and function of moving organs. However, direct application of a segmentation algorithm to each time frame of a sequence may ignore the temporal continuity inherent in the sequence. In this work, we propose an image sequence segmentation algorithm by combining a fully convolutional network with a recurrent neural network, which incorporates both spatial and temporal information into the segmentation task. A key challenge in training this network is that the available manual annotations are temporally sparse, which forbids end-to-end training. We address this challenge by performing non-rigid label propagation on the annotations and introducing an exponentially weighted loss function for training. Experiments on aortic MR image sequences demonstrate that the proposed method significantly improves both accuracy and temporal smoothness of segmentation, compared to a baseline method that utilises spatial information only. It achieves an average Dice metric of 0.960 for the ascending aorta and 0.953 for the descending aorta.
UR - http://www.scopus.com/inward/record.url?scp=85053834602&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85053834602&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-00937-3_67
DO - 10.1007/978-3-030-00937-3_67
M3 - Conference contribution
AN - SCOPUS:85053834602
SN - 9783030009366
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 586
EP - 594
BT - Medical Image Computing and Computer Assisted Intervention – MICCAI 2018 - 21st International Conference, 2018, Proceedings
A2 - Frangi, Alejandro F.
A2 - Fichtinger, Gabor
A2 - Schnabel, Julia A.
A2 - Alberola-López, Carlos
A2 - Davatzikos, Christos
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 -