TY - JOUR
T1 - 3deecelltracker, a deep learning-based pipeline for segmenting and tracking cells in 3d time lapse images
AU - Wen, Chentao
AU - Miura, Takuya
AU - Voleti, Venkatakaushik
AU - Yamaguchi, Kazushi
AU - Tsutsumi, Motosuke
AU - Yamamoto, Kei
AU - Otomo, Kohei
AU - Fujie, Yukako
AU - Teramoto, Takayuki
AU - Ishihara, Takeshi
AU - Aoki, Kazuhiro
AU - Nemoto, Tomomi
AU - Hillman, Elizabeth M.C.
AU - Kimura, Koutarou D.
N1 - Funding Information:
We thank Nobutoshi Odajima (Flovel Co. Ltd.), Hideki Tanaka (Yokogawa Corp.), Kenichi Matsumoto (Hamamatsu Photonics KK), and Kazuma Etani (Nikon Solutions Co. Ltd.) for setting up the OSB-3D. We also thank William Graf and David Van Valen for their kind advice in testing DeepCell 2.0, and Yu Toyoshima for his kind advice and help in testing the software by Toyoshima et al., 2016. We also thank Toru Tamaki, Ichiro Takeuchi, Takuto Sakuma, Katsuyoshi Matsushita, Hiroyuki Kaneko, Taro Sakurai, Jared Young, Richard Yan, Yuto Endo and the other Kimura laboratory members for their valuable advice, comments and technical assistance for this study. Nematode strains were provided by the Caenorhabditis Genetics Center (funded by the NIH Office of Research Infrastructure Programs P40 OD010440). Zebrafish samples were provided by Kimara Targoff, Caitlin Ford and Carmen de Sena Tomás and imaged with assistance from Citlali Perez-Campos and Wenze Li.
Funding Information:
We thank Nobutoshi Odajima (Flovel Co. Ltd.), Hideki Tanaka (Yokogawa Corp.), Kenichi Matsumoto (Hamamatsu Photonics KK), and Kazuma Etani (Nikon Solutions Co. Ltd.) for setting up the OSB-3D. We also thank William Graf and David Van Valen for their kind advice in testing DeepCell 2.0, and Yu Toyoshima for his kind advice and help in testing the software by Toyoshima et al., 2016. We also thank Toru Tamaki, Ichiro Takeuchi, Takuto Sakuma, Katsuyoshi Matsushita, Hiroyuki Kaneko, Taro Sakurai, Jared Young, Richard Yan, Yuto Endo and the other Kimura laboratory members for their valuable advice, comments and technical assistance for this study. Nematode strains were provided by the Caenorhabditis Genetics Center (funded by the NIH Office of Research Infrastructure Programs P40 OD010440). Zebrafish samples were provided by Kimara Targoff, Caitlin Ford and Carmen de Sena Tom?s and imaged with assistance from Citlali Perez-Campos and Wenze Li. Japan Society for the Promotion of Science- KAKENHI JP16H06545- Koutarou D Kimura; Japan Society for the Promotion of Science- KAKENHI JP20H05700- Koutarou D Kimura; Japan Society for the Promotion of Science- KAKENHI JP18H05135- Takeshi Ishihara; Japan Society for the Promotion of Science- KAKENHI JP19K15406- Motosuke Tsutsumi; NIH/NINDS- U01NS094296 UF1NS108213- Elizabeth MC Hillman; NIH/NCI- U01CA236554- Elizabeth MC Hillman; National Institutes of Natural Sciences- 01112002- Koutarou D Kimura; Grant-in-Aid for Research in Nagoya City University- 48 1912011 1921102- Koutarou D Kimura; RIKEN Center for Advanced Intelligence Project- Koutarou D Kimura; A program for Leading Graduate Schools entitled ?Interdisciplinary graduate school program for systematic understanding of health and disease?- Takuya Miura; NTT-Kyushu University Collaborative Research Program on Basic Science- Takeshi Ishihara.
Publisher Copyright:
© Wen et al.
PY - 2021/3
Y1 - 2021/3
N2 - Despite recent improvements in microscope technologies, segmenting and tracking cells in three-dimensional time-lapse images (3D + T images) to extract their dynamic positions and activities remains a considerable bottleneck in the field. We developed a deep learning-based software pipeline, 3DeeCellTracker, by integrating multiple existing and new techniques including deep learning for tracking. With only one volume of training data, one initial correction, and a few parameter changes, 3DeeCellTracker successfully segmented and tracked ~100 cells in both semi-immobilized and ‘straightened’ freely moving worm’s brain, in a naturally beating zebrafish heart, and ~1000 cells in a 3D cultured tumor spheroid. While these datasets were imaged with highly divergent optical systems, our method tracked 90–100% of the cells in most cases, which is comparable or superior to previous results. These results suggest that 3DeeCellTracker could pave the way for revealing dynamic cell activities in image datasets that have been difficult to analyze.
AB - Despite recent improvements in microscope technologies, segmenting and tracking cells in three-dimensional time-lapse images (3D + T images) to extract their dynamic positions and activities remains a considerable bottleneck in the field. We developed a deep learning-based software pipeline, 3DeeCellTracker, by integrating multiple existing and new techniques including deep learning for tracking. With only one volume of training data, one initial correction, and a few parameter changes, 3DeeCellTracker successfully segmented and tracked ~100 cells in both semi-immobilized and ‘straightened’ freely moving worm’s brain, in a naturally beating zebrafish heart, and ~1000 cells in a 3D cultured tumor spheroid. While these datasets were imaged with highly divergent optical systems, our method tracked 90–100% of the cells in most cases, which is comparable or superior to previous results. These results suggest that 3DeeCellTracker could pave the way for revealing dynamic cell activities in image datasets that have been difficult to analyze.
UR - http://www.scopus.com/inward/record.url?scp=85103316203&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85103316203&partnerID=8YFLogxK
U2 - 10.7554/eLife.59187
DO - 10.7554/eLife.59187
M3 - Article
C2 - 33781383
AN - SCOPUS:85103316203
VL - 10
JO - eLife
JF - eLife
SN - 2050-084X
M1 - e59187
ER -