Background Plant root research can provide a way to attain stress-tolerant crops that produce greater yield in a diverse array of conditions. Phenotyping roots in soil is often challenging due to the roots being difficult to access and the use of time consuming manual methods. Rhizotrons allow visual inspection of root growth through transparent surfaces. Agronomists currently manually label photographs of roots obtained from rhizotrons using a line-intersect method to obtain root length density and rooting depth measurements which are essential for their experiments. We investigate the effectiveness of an automated image segmentation method based on the U-Net Convolutional Neural Network (CNN) architecture to enable such measurements. We design a data-set of 50 annotated chicory (Cichorium intybus L.) root images which we use to train, validate and test the system and compare against a baseline built using the Frangi vesselness filter. We obtain metrics using manual annotations and line-intersect counts. Results Our results on the held out data show our proposed automated segmentation system to be a viable solution for detecting and quantifying roots. We evaluate our system using 867 images for which we have obtained line-intersect counts, attaining a Spearman rank correlation of 0.9748 and an $$r^2$$r2 of 0.9217. We also achieve an $$F_1$$F1 of 0.7 when comparing the automated segmentation to the manual annotations, with our automated segmentation system producing segmentations with higher quality than the manual annotations for large portions of the image. Conclusion We have demonstrated the feasibility of a U-Net based CNN system for segmenting images of roots in soil and for replacing the manual line-intersect method. The success of our approach is also a demonstration of the feasibility of deep learning in practice for small research groups needing to create their own custom labelled dataset from scratch.

We present RootPainter, a GUI-based software tool for the rapid training of deep neural networks for use in biological image analysis. RootPainter facilitates both fully-automatic and semiautomatic image segmentation. We investigate the effectiveness of RootPainter using three plant image datasets, evaluating its potential for root length extraction from chicory roots in soil, biopore counting and root nodule counting from scanned roots. We also use RootPainter to compare dense annotations to corrective ones which are added during the training based on the weaknesses of the current model. Deep Learning | GUI | Segmentation | Phenotyping | Biopore | Rhizotron | Root nodule | Interactive segmentation Correspondence: ags@di.ku.dk Fig. 1. RootPainter corrective annotation concept. (a) Roots in soil. (b) AI root predictions. (c) Human corrections. (d) AI learns from corrections. Smith et al. | bioRχiv | April 16, 2020 | 1-16

Patients with severe COVID-19 have overwhelmed healthcare systems worldwide. We hypothesized that machine learning (ML) models could be used to predict risks at different stages of management and thereby provide insights into drivers and prognostic markers of disease progression and death. From a cohort of approx. 2.6 million citizens in Denmark, SARS-CoV-2 PCR tests were performed on subjects suspected for COVID-19 disease; 3944 cases had at least one positive test and were subjected to further analysis. SARS-CoV-2 positive cases from the United Kingdom Biobank was used for external validation. The ML models predicted the risk of death (Receiver Operation Characteristics—Area Under the Curve, ROC-AUC) of 0.906 at diagnosis, 0.818, at hospital admission and 0.721 at Intensive Care Unit (ICU) admission. Similar metrics were achieved for predicted risks of hospital and ICU admission and use of mechanical ventilation. Common risk factors, included age, body mass index and hypertension, although the top risk features shifted towards markers of shock and organ dysfunction in ICU patients. The external validation indicated fair predictive performance for mortality prediction, but suboptimal performance for predicting ICU admission. ML may be used to identify drivers of progression to more severe disease and for prognostication patients in patients with COVID-19. We provide access to an online risk calculator based on these findings.

We present RootPainter, a GUI-based software tool for the rapid training of deep neural networks for use in biological image analysis. RootPainter facilitates both fully-automatic and semiautomatic image segmentation. We investigate the effectiveness of RootPainter using three plant image datasets, evaluating its potential for root length extraction from chicory roots in soil, biopore counting and root nodule counting from scanned roots. We also use RootPainter to compare dense annotations to corrective ones which are added during the training based on the weaknesses of the current model.

Deep rooting winter wheat genotypes can reduce nitrate leaching losses and increase N uptake. We aimed to investigate which deep root traits are correlated to deep N uptake and to estimate genetic variation in root traits and deep 15 N tracer uptake. In 2 years, winter wheat genotypes were grown in RadiMax, a semifield root-screening facility.Minirhizotron root imaging was performed three times during the main growing season.At anthesis, 15 N was injected via subsurface drip irrigation at 1.8 m depth. Mature ears from above the injection area were analysed for 15 N content. From minirhizotron imagebased root length data, 82 traits were constructed, describing root depth, density, distribution and growth aspects. Their ability to predict 15 N uptake was analysed with the least absolute shrinkage and selection operator (LASSO) regression. Root traits predicted 24% and 14% of tracer uptake variation in 2 years. Both root traits and genotype showed significant effects on tracer uptake. In 2018, genotype and the three LASSO-selected root traits predicted 41% of the variation in tracer uptake, in 2019 genotype and one root trait predicted 48%. In both years, one root trait significantly mediated the genotype effect on tracer uptake. Deep root traits from minirhizotron images can predict deep N uptake, indicating the potential to breed deep-N-uptake-genotypes.nitrogen isotope, plant breeding, plant roots, semifield | INTRODUCTIONDeeper rooting crops expand the soil depth from which nitrogen (N) can be taken up. This increases the N use efficiency of cropping systems and decreases leaching losses (Dresbøll & Thorup-Kristensen, 2014). The primary form of mineral N in temperate soils is nitrate, which is highly mobile in the soil water solution, as it is a negatively charged molecule (Allred et al., 2007). Therefore, nitrate percolates with excess precipitation and reaches deep soil layers easily. If subsequent crop roots do not penetrate to these soil layers, nitrate leached so deep that it is close to the bottom of the root zone will be lost from the cropping system. If we expand the rooting depth of crops, we can increase the uptake of leached nitrate (e.g., Thorup-Kristensen, 2006) and increase total crop N uptake (Thorup-Kristensen et al., 2009). The importance of deep roots for N acquisition in a leaching situation was suggested in the 'steep-cheap-deep' ideotype concept (Lynch, 2013).A 2-year Danish field experiment found significant variance in root depth between cultivars and that deeper roots extracted more N from deep soil (Rasmussen et al., 2015). These findings are supported by other studies that measured the deep root N uptake of crops through 15 N injection into deep soil layers (Chen et al., 2019;Saengwilai et al., 2014;Kristensen & Thorup-Kristensen, 2004a, 2004b. These studies found significant variation between species (Kristensen & Thorup-Kristensen, 2004a, 2004b) and within species (Chen

Background: Patients with severe COVID-19 have overwhelmed healthcare systems worldwide. We hypothesized that Machine Learning (ML) models could be used to predict risks at different stages of management (at diagnosis, hospital admission and ICU admission) and thereby provide insights into drivers and prognostic markers of disease progression and death. Methods: From a cohort of approx. 2.6 million citizens in the two regions of Denmark, SARS-CoV-2 PCR tests were performed on subjects suspected for COVID-19 disease; 3944 cases had at least one positive test and were subjected to further analysis. A cohort of SARS-CoV-2 positive cases from the United Kingdom Biobank was used for external validation. Findings: The ML models predicted the risk of death (Receiver Operation Characteristics Area Under the Curve, ROC-AUC) of 0.904 at diagnosis, 0.818, at hospital admission and 0.723 at Intensive Care Unit (ICU) admission. Similar metrics were achieved for predicted risks of hospital and ICU admission and use of mechanical ventilation. We identified some common risk factors, including age, body mass index (BMI) and hypertension as driving factors, although the top risk features shifted towards markers of shock and organ dysfunction in ICU patients. The external validation indicated fair predictive performance for mortality prediction, but suboptimal performance for predicting ICU admission. Interpretation: ML may be used to identify drivers of progression to more severe disease and for prognostication patients in patients with COVID-19. Prognostic features included age, BMI and hypertension, although markers of shock and organ dysfunction became more important in more severe cases. We provide access to an online risk calculator based on these findings.

The scale of root quantification in research is often limited by the time required for sampling, measurement and processing samples. Recent developments in Convolutional Neural Networks (CNN) have made faster and more accurate plant image analysis possible which may significantly reduce the time required for root measurement, but challenges remain in making these methods accessible to researchers without an in-depth knowledge of Machine Learning. We analyzed root images acquired from three destructive root samplings using the RootPainter CNN-software that features an interface for corrective annotation for easier use. Root scans with and without non-root debris were used to test if training a model, i.e., learning from labeled examples, can effectively exclude the debris by comparing the end-results with measurements from clean images. Root images acquired from soil profile walls and the cross-section of soil cores were also used for training and the derived measurements were compared with manual measurements. After 200 minutes of training on each dataset, significant relationships between manual measurements and RootPainter-derived data were noted for monolith (R2=0.99), profile wall (R2=0.76) and core-break (R2=0.57). The rooting density derived from images with debris was not significantly different from that derived from clean images after processing with RootPainter. Rooting density was also successfully calculated from both profile wall and soil core images, and in each case the gradient of root density with depth was not significantly different from manual counts. Our results demonstrate that the proposed approach using CNN can lead to substantial reductions in root sample processing workloads, increasing the potential scale of future root investigations.

The scale of root quantification in research is often limited by the time required for sampling, measurement, and processing samples. Recent developments in convolutional neural networks (CNNs) have made faster and more accurate plant image analysis possible, which may significantly reduce the time required for root measurement, but challenges remain in making these methods accessible to researchers without an in-depth knowledge of machine learning. We analyzed root images acquired from three destructive root samplings using the RootPainter CNN software that features an interface for corrective annotation for easier use. Root scans with and without non-root debris were used to test if training a model (i.e. learning from labeled examples) can effectively exclude the debris by comparing the end results with measurements from clean images. Root images acquired from soil profile walls and the cross-section of soil cores were also used for training, and the derived measurements were compared with manual measurements. After 200 min of training on each dataset, significant relationships between manual measurements and RootPainter-derived data were noted for monolith (R2=0.99), profile wall (R2=0.76), and core-break (R2=0.57). The rooting density derived from images with debris was not significantly different from that derived from clean images after processing with RootPainter. Rooting density was also successfully calculated from both profile wall and soil core images, and in each case the gradient of root density with depth was not significantly different from manual counts. Differences in root-length density (RLD) between crops with contrasting root systems were captured using automatic segmentation at soil profiles with high RLD (1–5 cm cm–3) as well with low RLD (0.1–0.3 cm cm–3). Our results demonstrate that the proposed approach using CNN can lead to substantial reductions in root sample processing workloads, increasing the potential scale of future root investigations.