A deep learning-based approach for high-throughput hypocotyl phenotyping

Dobos, Orsolya; Horváth, Péter; Nagy, Ferenc; Danka, Tivadar; Viczián, András

doi:10.1101/651729

Cited by 8 publications

(8 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Numerous approaches have been developed, including methods based on mathematical morphology 1 or differential geometry 2,3 . More recently, deep learning has yielded a never-seen improvement of accuracy and robustness [4][5][6] . Remarkably, Kaggle's Data Science Bowl 2018 (DSB) 7 was dedicated to nuclei segmentation, and gave a great momentum to this field.…”

mentioning

confidence: 99%

Test-time augmentation for deep learning-based cell segmentation on microscopy images

Moshkov

Mathe

Kertész‐Farkas

et al. 2020

Sci Rep

Self Cite

159

View full text Add to dashboard Cite

Recent advancements in deep learning have revolutionized the way microscopy images of cells are processed. Deep learning network architectures have a large number of parameters, thus, in order to reach high accuracy, they require a massive amount of annotated data. A common way of improving accuracy builds on the artificial increase of the training set by using different augmentation techniques. A less common way relies on test-time augmentation (TTA) which yields transformed versions of theimage for prediction and the results are merged. in this paper we describe how we have incorporated the test-time argumentation prediction method into two major segmentation approaches utilized in the single-cell analysis of microscopy images. these approaches are semantic segmentation based on the U-Net, and instance segmentation based on the Mask R-CNN models. Our findings show that even if only simple test-time augmentations (such as rotation or flipping and proper merging methods) are applied, TTA can significantly improve prediction accuracy. We have utilized images of tissue and cell cultures from the Data Science Bowl (DSB) 2018 nuclei segmentation competition and other sources. Additionally, boosting the highest-scoring method of the DSB with TTA, we could further improve prediction accuracy, and our method has reached an ever-best score at the DSB.

show abstract

mentioning

confidence: 99%

Test-time augmentation for deep learning-based cell segmentation on microscopy images

Moshkov

Mathe

Kertész‐Farkas

et al. 2020

Sci Rep

Self Cite

159

View full text Add to dashboard Cite

show abstract

“…Compared with other algorithms that use deep learning for image segmentation, the PlantU-net model can improve the segmentation precision by 10% compared with the U-net model [37] (Table 2), indicating that the PlantU-net model has higher credibility in the application of top-view images segmentation of maize plants at the seedling stage. The method proposed by Orsolya Dobos et al [45] uses U-net and 2,850 images to train the Arabidopsis image segmentation model, while the PlantU-net model only needs 512 images for training and the training data does not need complex pre-processing, indicating that PlantU-net achieves high-precision segmentation with less training data. Therefore, when PlantU-net is used to solve image segmentation problems in other crops at the seedling stage, only a small number of annotated images are needed, indicating that the method is highly scalable.…”

Section: Discussionmentioning

confidence: 99%

High-throughput phenotyping analysis of maize at the seedling stage using end-to-end segmentation network

Wen

Guo

et al. 2020

Preprint

View full text Add to dashboard Cite

Image processing technologies are available for high-throughput acquisition and analysis of phenotypes for crop populations, which is of great significance for crop growth monitoring, evaluation of seedling condition, and cultivation management. However, existing methods rely on empirical segmentation thresholds, thus can have insufficient accuracy of extracted phenotypes. Taking maize as an example crop, we propose a phenotype extraction approach from top-view images at the seedling stage. An end-to-end segmentation network, named PlantU-net, which uses a small amount of training data, was explored to realize automatic segmentation of top-view images of a maize population at the seedling stage. Morphological and color related phenotypes were automatic extracted, including maize shoot coverage, circumscribed radius, aspect ratio, and plant azimuth plane angle. The results show that the approach can segment the shoots at the seedling stage from top-view images, obtained either from the UAV or ground high-throughput phenotyping platform. The average segmentation accuracy, recall rate, and F1 score are 0.96, 0.98, and 0.97, respectively. The extracted phenotypes, including maize shoot coverage, circumscribed radius, aspect ratio, and plant azimuth plane angle, are highly correlated with manual measurements (R 2 =0.96-0.99). This approach requires less training data and thus has better expansibility. It provides practical means for high-throughput phenotyping analysis of early growth stage crop populations.

show abstract

“…Measurement of hypocotyl elongation and data evaluation were performed as described previously (Adam et al ., 2013; Dobos et al ., 2019). In the survival test, seedlings were grown in the dark on Murashige & Skoog (MS) medium plates for 4 d and were then placed under white light irradiation (100 μmol m −2 s −1 ) for 2 d. The ratio of total to survived seedlings was calculated.…”

Section: Methodsmentioning

confidence: 99%

SUMOylation of PHYTOCHROME INTERACTING FACTOR 3 promotes photomorphogenesis in Arabidopsis thaliana

et al. 2020

Self Cite

View full text Add to dashboard Cite

In Arabidopsis thaliana, phytochrome B (phyB) is the dominant receptor of photomorphogenic development under red light. Phytochrome B interacts with a set of downstream regulatory proteins, including PHYTOCHROME INTERACTING FACTOR 3 (PIF3). The interaction between PIF3 and photoactivated phyB leads to the rapid phosphorylation and degradation of PIF3 and also to the degradation of phyB, events which are required for proper photomorphogenesis. Here we report that PIF3 is SUMOylated at the Lys13 (K13) residue and that we could detect this posttranslational modification in a heterologous experimental system and also in planta. We also found that the SUMO acceptor site mutant PIF3(K13R) binds more strongly to the target promoters than its SUMOylated, wild-type counterpart. Seedlings expressing PIF3 (K13R) show an elongated hypocotyl response, elevated photoprotection and higher transcriptional induction of red-light responsive genes compared with plantlets expressing wildtype PIF3. These observations are supported by the lower level of phyB in plants which possess only PIF3(K13R), indicating that SUMOylation of PIF3 also alters photomorphogenesis via the regulation of phyB levels. In conclusion, whereas SUMOylation is generally connected to different stress responses, it also fine-tunes light signalling by reducing the biological activity of PIF3, thus promoting photomorphogenesis.

show abstract

A deep learning-based approach for high-throughput hypocotyl phenotyping

Cited by 8 publications

References 49 publications

Test-time augmentation for deep learning-based cell segmentation on microscopy images

Test-time augmentation for deep learning-based cell segmentation on microscopy images

High-throughput phenotyping analysis of maize at the seedling stage using end-to-end segmentation network

SUMOylation of PHYTOCHROME INTERACTING FACTOR 3 promotes photomorphogenesis in Arabidopsis thaliana

Contact Info

Product

Resources

About