2018年3月18日
Cross-modality image synthesis from unpaired data using CycleGAN: Effects of gradient consistency loss and training data size
SIMULATION AND SYNTHESIS IN MEDICAL IMAGING
- ,
- ,
- ,
- ,
- ,
- ,
- ,
- 巻
- 11037
- 号
- 開始ページ
- 31
- 終了ページ
- 41
- 記述言語
- 掲載種別
- 研究論文(国際会議プロシーディングス)
- DOI
- 10.1007/978-3-030-00536-8_4
- 出版者・発行元
- Springer
CT is commonly used in orthopedic procedures. MRI is used along with CT to
identify muscle structures and diagnose osteonecrosis due to its superior soft
tissue contrast. However, MRI has poor contrast for bone structures. Clearly,
it would be helpful if a corresponding CT were available, as bone boundaries
are more clearly seen and CT has standardized (i.e., Hounsfield) units.
Therefore, we aim at MR-to-CT synthesis. The CycleGAN was successfully applied
to unpaired CT and MR images of the head, these images do not have as much
variation of intensity pairs as do images in the pelvic region due to the
presence of joints and muscles. In this paper, we extended the CycleGAN
approach by adding the gradient consistency loss to improve the accuracy at the
boundaries. We conducted two experiments. To evaluate image synthesis, we
investigated dependency of image synthesis accuracy on 1) the number of
training data and 2) the gradient consistency loss. To demonstrate the
applicability of our method, we also investigated a segmentation accuracy on
synthesized images.
identify muscle structures and diagnose osteonecrosis due to its superior soft
tissue contrast. However, MRI has poor contrast for bone structures. Clearly,
it would be helpful if a corresponding CT were available, as bone boundaries
are more clearly seen and CT has standardized (i.e., Hounsfield) units.
Therefore, we aim at MR-to-CT synthesis. The CycleGAN was successfully applied
to unpaired CT and MR images of the head, these images do not have as much
variation of intensity pairs as do images in the pelvic region due to the
presence of joints and muscles. In this paper, we extended the CycleGAN
approach by adding the gradient consistency loss to improve the accuracy at the
boundaries. We conducted two experiments. To evaluate image synthesis, we
investigated dependency of image synthesis accuracy on 1) the number of
training data and 2) the gradient consistency loss. To demonstrate the
applicability of our method, we also investigated a segmentation accuracy on
synthesized images.
- リンク情報
-
- DOI
- https://doi.org/10.1007/978-3-030-00536-8_4
- DBLP
- https://dblp.uni-trier.de/rec/conf/miccai/HiasaOTMTCPSS18
- arXiv
- http://arxiv.org/abs/arXiv:1803.06629
- Web of Science
- https://gateway.webofknowledge.com/gateway/Gateway.cgi?GWVersion=2&SrcAuth=JSTA_CEL&SrcApp=J_Gate_JST&DestLinkType=FullRecord&KeyUT=WOS:000477752900004&DestApp=WOS_CPL
- URL
- http://arxiv.org/abs/1803.06629v3
- URL
- http://arxiv.org/pdf/1803.06629v3 本文へのリンクあり
- Scopus
- https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85053912661&origin=inward
- Scopus Citedby
- https://www.scopus.com/inward/citedby.uri?partnerID=HzOxMe3b&scp=85053912661&origin=inward
- ID情報
-
- DOI : 10.1007/978-3-030-00536-8_4
- ISSN : 0302-9743
- eISSN : 1611-3349
- ISBN : 9783030005351
- DBLP ID : conf/miccai/HiasaOTMTCPSS18
- arXiv ID : arXiv:1803.06629
- SCOPUS ID : 85053912661
- Web of Science ID : WOS:000477752900004