deep learning-based shoot image analysis tool deepshoot  (LemnaTec Inc)

Journal: Frontiers in Plant Science

Article Title: Deep Learning Based Greenhouse Image Segmentation and Shoot Phenotyping (DeepShoot)

doi: 10.3389/fpls.2022.906410

Figure Lengend Snippet: Workflow of the pipeline for image processing and segmentation in the DeepShoot tool. Green and orange color boxes represent the operations of image segmentation and trait calculation: (A) original image, (B) original image patches of size 256 x 256, (C) segmented image patches of size 256 x 256, (D) binary segmentation of the original image, (E) RGB color space of (D) .

Article Snippet: The deep learning-based shoot image analysis tool (DeepShoot) is designed for automated segmentation and quantification of visible light (VIS) images of arabidopsis, maize, and barley shoots acquired from greenhouse phenotyping experiments using LemnaTec-Scanalyzer3D high throughput plant phenotyping platforms (LemnaTec GmbH, Aachen, Germany). shows examples of arabidopsis, maize, and barely images from three different LemnaTec phenotyping platforms tailored to the screening of large, mid-size, and small plants.

Techniques:

Journal: Frontiers in Plant Science

Article Title: Deep Learning Based Greenhouse Image Segmentation and Shoot Phenotyping (DeepShoot)

doi: 10.3389/fpls.2022.906410

Figure Lengend Snippet: Segmentation performance: first, second and third row represents the original RGB image, ground truth segmentation by the kmSeg tool and predicted segmentation by the DeepShoot tool, respectively. The DC of each image as following: (A) Arabidopsis side view: 0.9117, (B) Arabidopsis top view: 0.9876, (C) Barley side view: 0.9384, (D) Barley top view: 0.9617, (E) Maize side view: 0.9709, (F) Maize top view: 0.9843.

Article Snippet: The deep learning-based shoot image analysis tool (DeepShoot) is designed for automated segmentation and quantification of visible light (VIS) images of arabidopsis, maize, and barley shoots acquired from greenhouse phenotyping experiments using LemnaTec-Scanalyzer3D high throughput plant phenotyping platforms (LemnaTec GmbH, Aachen, Germany). shows examples of arabidopsis, maize, and barely images from three different LemnaTec phenotyping platforms tailored to the screening of large, mid-size, and small plants.

Techniques:

Journal: Frontiers in Plant Science

Article Title: Deep Learning Based Greenhouse Image Segmentation and Shoot Phenotyping (DeepShoot)

doi: 10.3389/fpls.2022.906410

Figure Lengend Snippet: Evaluation of neural network segmentation with respect to the proposed U-net on arabidopsis, barley and maize side view images respectively. (A) NN DC: 0.7824, DeepShoot DC: 0.8342, (B) NN DC: 0.6973, DeepShoot DC: 0.8924, (C) NN DC: 0.8746, DeepShoot DC: 0.9360.

Article Snippet: The deep learning-based shoot image analysis tool (DeepShoot) is designed for automated segmentation and quantification of visible light (VIS) images of arabidopsis, maize, and barley shoots acquired from greenhouse phenotyping experiments using LemnaTec-Scanalyzer3D high throughput plant phenotyping platforms (LemnaTec GmbH, Aachen, Germany). shows examples of arabidopsis, maize, and barely images from three different LemnaTec phenotyping platforms tailored to the screening of large, mid-size, and small plants.

Techniques:

Journal: Frontiers in Plant Science

Article Title: Deep Learning Based Greenhouse Image Segmentation and Shoot Phenotyping (DeepShoot)

doi: 10.3389/fpls.2022.906410

Figure Lengend Snippet: Graphical User Interface of the DeepShoot tool: left, middle, and right images represent the original image, predicted probability map, and predicted color image, respectively.

Article Snippet: The deep learning-based shoot image analysis tool (DeepShoot) is designed for automated segmentation and quantification of visible light (VIS) images of arabidopsis, maize, and barley shoots acquired from greenhouse phenotyping experiments using LemnaTec-Scanalyzer3D high throughput plant phenotyping platforms (LemnaTec GmbH, Aachen, Germany). shows examples of arabidopsis, maize, and barely images from three different LemnaTec phenotyping platforms tailored to the screening of large, mid-size, and small plants.

Techniques:

Journal: Scientific Reports

Article Title: Identification of an elusive spliceogenic MYBPC3 variant in an otherwise genotype-negative hypertrophic cardiomyopathy pedigree

doi: 10.1038/s41598-022-11159-y

Figure Lengend Snippet: Frequency of cryptic splice-altering MYBPC3 variants in a large cohort of HCM probands.

Article Snippet: The probability of a variant being splice-altering was estimated in silico using the deep learning-based tool SpliceAI with a window of ± 200 bp (maximum distance of analyzed nucleotides from the variant) unless otherwise indicated, and the splicing prediction module of the software package Alamut Visual (v.2.11.0, Interactive Biosoftware).

Techniques: