EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Deep learning-based high-throughput detection of in vitro germination to assess pollen viability from microscopic images

Mengwei Zhang, Jianxiang Zhao, Yoichiro Hoshino iD

Journal of Experimental Botany · 21.11.2023 · 74 (21) · 6551-6562

DOI: 10.1093/jxb/erad315 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Experimental Botany, online erschienen am 16.08.2023. Verfasst von Mengwei Zhang und 2 weiteren Autor:innen. Beteiligte Einrichtung: Division of Biosphere Science, Graduate School of Environmental Science, Hokkaido University, Kita 11, Nishi 10, Kita-ku, Sapporo 060-0811, Japan. Hauptthemen laut MEDLINE: Germination, Deep Learning. Die Arbeit wurde bisher 14-mal zitiert (Crossref) und nennt 59 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Japan Science and Technology Agency, Next Generation.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

In vitro pollen germination is considered the most efficient method to assess pollen viability. The pollen germination frequency and pollen tube length, which are key indicators of pollen viability, should be accurately measured during in vitro culture. In this study, a Mask R-CNN model trained using microscopic images of tree peony (Paeonia suffruticosa) pollen has been proposed to rapidly detect the pollen germination rate and pollen tube length. To reduce the workload during image acquisition, images of synthesized crossed pollen tubes were added to the training dataset, significantly improving the model accuracy in recognizing crossed pollen tubes. At an Intersection over Union threshold of 50%, a mean average precision of 0.949 was achieved. The performance of the model was verified using 120 testing images. The R2 value of the linear regression model using detected pollen germination frequency against the ground truth was 0.909 and that using average pollen tube length was 0.958. Further, the model was successfully applied to two other plant species, indicating a good generalizability and potential to be applied widely.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

GerminationDeep LearningPollenPollen Tube

Stichwörter der Autor:innen

Paeonia suffruticosaDeep learningMask R-CNNpollen germination frequencypollen tube lengthpollen viability

Autor:innen und Institutionen

  • Division of Biosphere Science, Graduate School of Environmental Science, Hokkaido University, Kita 11, Nishi 10, Kita-ku, Sapporo 060-0811, Japan.
  • Field Science Center for Northern Biosphere, Hokkaido University, Kita 11, Nishi 10, Kita-ku, Sapporo 060-0811, Japan.

ORCID-Profile: Yoichiro Hoshino

Förderung

  • Japan Science and Technology Agency
  • Next Generation

Zitieren

APA

Zhang, M., Zhao, J., & Hoshino, Y. (2023). Deep learning-based high-throughput detection of in vitro germination to assess pollen viability from microscopic images. Journal of Experimental Botany, 74(21), 6551-6562. https://doi.org/10.1093/jxb/erad315

Vancouver

Zhang M, Zhao J, Hoshino Y. Deep learning-based high-throughput detection of in vitro germination to assess pollen viability from microscopic images. Journal of Experimental Botany. 2023;74(21):6551-6562. doi:10.1093/jxb/erad315

BibTeX
@article{Zhang2023,
  author = {Zhang, Mengwei and Zhao, Jianxiang and Hoshino, Yoichiro},
  title = {Deep learning-based high-throughput detection of in vitro germination to assess pollen viability from microscopic images},
  journal = {Journal of Experimental Botany},
  year = {2023},
  volume = {74},
  number = {21},
  pages = {6551--6562},
  doi = {10.1093/jxb/erad315}
}
Kostenlos starten →