Presentation 2019-01-23
Unsupervised Shadow Detection for Ultrasound Images by Deep Learning
Suguru Yasutomi, Akira Sakai, Masaaki Komatsu, Ryu Matsuoka, Reina Komatsu, Tatsuya Arakaki, Mayumi Tokunaka, Hidenori Machino, Kazuma Kobayashi, Ken Asada, Syuzo Kaneko, Akihiko Sekizawa, Ryuji Hamamoto,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Medical ultrasound is widely used for diagnosing internal organs since it is non-invasive. Shadows are often appear in ultrasound images, and they hinder diagnosing and image processing. Detecting shadows from the images is important problem in such cases. In this paper, we propose a method to detect shadows using deep learning which is learned by unsupervised way. Specifically, we construct an autoencoder that splits input images into shadows and other contents, and combines them to reconstruct the inputs. The autoencoder is learned to split by newly proposed losses that evaluates characteristics of ultrasound images and its shadows. Effectiveness of the proposed method is shown by experiments on images for embryonic heart diagnosis.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Ultrasound imaging / Shadow detection / Deep learning / Unsupervised learning
Paper # MI2018-96
Date of Issue 2019-01-15 (MI)

Conference Information
Committee MI
Conference Date 2019/1/22(2days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English) Medical Image Engineering, Analysis, Recognition, etc.
Chair Kensaku Mori(Nagoya Univ.)
Vice Chair Yoshiki Kawata(Tokushima Univ.) / Yuichi Kimura(Kinki Univ.)
Secretary Yoshiki Kawata(Aichi Inst. of Tech.) / Yuichi Kimura(Nagoya Inst. of Tech.)
Assistant Ryo Haraguchi(Univ. of Hyogo) / Yasushi Hirano(Yamaguchi Univ.)

Paper Information
Registration To Medical Imaging
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Unsupervised Shadow Detection for Ultrasound Images by Deep Learning
Sub Title (in English)
Keyword(1) Ultrasound imaging
Keyword(2) Shadow detection
Keyword(3) Deep learning
Keyword(4) Unsupervised learning
1st Author's Name Suguru Yasutomi
1st Author's Affiliation Fujitsu Laboratories Ltd., Artificial Intelligence Laboratory(FLL)
2nd Author's Name Akira Sakai
2nd Author's Affiliation Fujitsu Advanced Technologies, Division of Development Platform Technology(FATEC)
3rd Author's Name Masaaki Komatsu
3rd Author's Affiliation RIKEN, AIP Center, Cancer Translational Research Team(Riken)
4th Author's Name Ryu Matsuoka
4th Author's Affiliation Showa University School of Medicine, Department of Obstetrics and Gynecology(Showa-U)
5th Author's Name Reina Komatsu
5th Author's Affiliation Showa University School of Medicine, Department of Obstetrics and Gynecology(Showa-U)
6th Author's Name Tatsuya Arakaki
6th Author's Affiliation Showa University School of Medicine, Department of Obstetrics and Gynecology(Showa-U)
7th Author's Name Mayumi Tokunaka
7th Author's Affiliation Showa University School of Medicine, Department of Obstetrics and Gynecology(Showa-U)
8th Author's Name Hidenori Machino
8th Author's Affiliation National Cancer Center Research Institute, Division of Molecular Modification and Cancer Biology(NCC)
9th Author's Name Kazuma Kobayashi
9th Author's Affiliation National Cancer Center Research Institute, Division of Molecular Modification and Cancer Biology(NCC)
10th Author's Name Ken Asada
10th Author's Affiliation RIKEN, AIP Center, Cancer Translational Research Team(Riken)
11th Author's Name Syuzo Kaneko
11th Author's Affiliation National Cancer Center Research Institute, Division of Molecular Modification and Cancer Biology(NCC)
12th Author's Name Akihiko Sekizawa
12th Author's Affiliation Showa University School of Medicine, Department of Obstetrics and Gynecology(Showa-U)
13th Author's Name Ryuji Hamamoto
13th Author's Affiliation RIKEN, AIP Center, Cancer Translational Research Team(Riken)
Date 2019-01-23
Paper # MI2018-96
Volume (vol) vol.118
Number (no) MI-412
Page pp.pp.151-156(MI),
#Pages 6
Date of Issue 2019-01-15 (MI)