A5006905756- Plant nutrient uptake and metabolism
- Smart Agriculture and AI
- Rice Cultivation and Yield Improvement
- Advanced Radiotherapy Techniques
- Data Mining Algorithms and Applications
- COVID-19 diagnosis using AI
- Tree Root and Stability Studies
- Genetic Mapping and Diversity in Plants and Animals
- Wheat and Barley Genetics and Pathology
- Coral and Marine Ecosystems Studies
- Bioenergy crop production and management
- COVID-19 Clinical Research Studies
- 3D Modeling in Geospatial Applications
- Effects of Radiation Exposure
- Plant Micronutrient Interactions and Effects
- Environmental DNA in Biodiversity Studies
- Advanced MRI Techniques and Applications
- Genomics and Phylogenetic Studies
- Irrigation Practices and Water Management
- Seedling growth and survival studies
- Crop Yield and Soil Fertility
- Medical Imaging Techniques and Applications
- Soil Moisture and Remote Sensing
- Radiation Dose and Imaging
- Soil Carbon and Nitrogen Dynamics
Aarhus University
2024
University of Copenhagen
2019-2024
Rigshospitalet
2021-2023
Copenhagen University Hospital
2022
Planta
2020
Summary Convolutional neural networks (CNNs) are a powerful tool for plant image analysis, but challenges remain in making them more accessible to researchers without machine‐learning background. We present R oot P ainter , an open‐source graphical user interface based software the rapid training of deep use biological analysis. evaluate by models root length extraction from chicory ( Cichorium intybus L.) roots soil, biopore counting, and nodule counting. also compare dense annotations with...
Abstract Background Plant root research can provide a way to attain stress-tolerant crops that produce greater yield in diverse array of conditions. Phenotyping roots soil is often challenging due the being difficult access and use time consuming manual methods. Rhizotrons allow visual inspection growth through transparent surfaces. Agronomists currently manually label photographs obtained from rhizotrons using line-intersect method obtain length density rooting depth measurements which are...
Abstract 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 prognostic markers disease progression death. From a cohort approx. 2.6 million citizens in Denmark, SARS-CoV-2 PCR tests were performed on subjects suspected for disease; 3944 cases had least one positive test subjected further analysis. from the...
We present RootPainter, a GUI-based software tool for the rapid training of deep neural networks use in biological image analysis. RootPainter facilitates both fully-automatic and semi-automatic segmentation. investigate effectiveness using three plant datasets, evaluating its potential root length extraction from chicory roots soil, biopore counting nodule scanned roots. also to compare dense annotations corrective ones which are added during based on weaknesses current model.
MR-guided radiotherapy (MRgRT) allows real-time beam-gating to compensate for intra-fractional target position variations. This study investigates the dosimetric impact of and PTV margin on prostate coverage cancer patients treated with online-adaptive MRgRT.20 consecutive were MRgRT SBRT 36.25 Gy in 5 fractions (PTV D95% ≥ 95% (N = 5) 100% 15)). Sagittal 2D cine MRIs used gating a 3 mm expansion as window. We computed motion-compensated dose distributions (i) all positions during treatment...
Abstract Crop root segmentation models developed through deep learning have increased the throughput of in situ crop phenotyping studies. However, trained to identify roots one image dataset may not accurately another dataset, especially when new contains known differences, called domain shifts. The objective this study was quantify how model performance changes are used segment datasets that contain shifts and evaluate approaches reduce error associated with We collected maize images at two...
Abstract The scale of root quantification in research is often limited by the time required for sampling, measurement, and processing samples. Recent developments convolutional neural networks (CNNs) have made faster more accurate plant image analysis possible, which may significantly reduce but challenges remain making these methods accessible to researchers without an in-depth knowledge machine learning. We analyzed images acquired from three destructive samplings using RootPainter CNN...
Abstract Background Manual analysis of (mini-)rhizotron (MR) images is tedious. Several methods have been proposed for semantic root segmentation based on homogeneous, single-source MR datasets. Recent advances in deep learning (DL) enabled automated feature extraction, but comparisons accuracy, false positives and transferability are virtually lacking. Here we compare six state-of-the-art propose two improved DL models using a large dataset with without augmented data. We determine the...
Advancing imaging technologies are drastically increasing the rate of marine video and image data collection. Often these datasets not analysed to their full potential as extracting information for multiple species is incredibly time-consuming. This study demonstrates capability open-source interactive machine learning tool, RootPainter, analyse large quickly accurately. The ability RootPainter extract presence surface area cold-water coral reef associate sponge species,
Measuring seminal root angle is an important aspect of phenotyping, yet automated methods are lacking. We introduce SeminalRootAngle, a novel open-source method that measures angles from images. To ensure our flexible and user-friendly we build on established corrective annotation training for image segmentation. tested SeminalRootAngle heterogeneous dataset 662 spring barley rhizobox images, which presented challenges in terms clarity obstruction. Validation new pipeline against manual...
Soil biopore genesis is a dynamic and context-dependent process. Yet integrative investigations of under varying soil type, tillage vegetation history are rare. Recent advances in Machine Learning (ML) made faster more accurate image analysis possible. We validated model trained on Convolutional Neural Network (CNN) using multisite dataset from types (Luvisol, Cambisol Kandosol), (deep ploughing without deep ploughing) (taprooted fibrous-rooted crops) to automatically predict formation. The...
Background and purposeDaily plan adaptations could take the dose delivered in previous fractions into account. Due to high per fraction, low number of fractions, steep gradients, large interfractional organ deformations, this might be particularly important for liver SBRT. This study investigates inter-algorithm variation accumulation MR-guided SBRT.Materials methodsWe assessed 27 consecutive SBRT treatments 67.5 Gy three (n = 15) or 50 five 12), both prescribed GTV. We calculated fraction...
Deep rooting winter wheat genotypes can reduce nitrate leaching losses and increase N uptake. We aimed to investigate which deep root traits are correlated uptake estimate genetic variation in 15 tracer In 2 years, were grown RadiMax, a semifield root-screening facility. Minirhizotron imaging was performed three times during the main growing season. At anthesis, injected via subsurface drip irrigation at 1.8 m depth. Mature ears from above injection area analysed for content. From...
Organ-at-risk contouring is still a bottleneck in radiotherapy, with many deep learning methods falling short of promised results when evaluated on clinical data. We investigate the accuracy and time-savings resulting from use an interactive-machine-learning method for organ-at-risk task.We implement open-source software application that facilitates corrective-annotation deep-learning generated contours X-ray CT images. A trained-physician contoured 933 hearts using our by delineating first...
Abstract Background and purpose In the last 20 years, it has become well-documented that incidental cardiac exposure to ionizing radiation is associated with a clinically relevant increased risk of cardiovascular morbidity. parallel, therapy technologies have been developed provide target dose coverage less adjacent organs at risk. current work, we investigate trends in among patients treated curative intent radiotherapy from single institution between 2009 2020. Materials methods 10,215...
Increased organ at risk segmentation accuracy is required to reduce cost and complications for patients receiving radiotherapy treatment. Some deep learning methods the of organs use a two stage process where localisation network first crops an image relevant region then locally specialised segments cropped interest. We investigate improvements brought about by such comparing single-stage baseline trained on full resolution images. find that approaches can improve both training time...
Abstract The growing demand for food and feed crops in the world because of population more extreme weather events requires high‐yielding resilient crops. Many agriculturally important traits are polygenic, controlled by multiple regulatory layers, with a strong interaction environment. In this study, 120 F 2 families perennial ryegrass ( Lolium perenne L.) were grown across water gradient semifield facility subsoil irrigation. Genomic (single‐nucleotide polymorphism [SNP]), transcriptomic...
Enhanced nitrogen (N) and water uptake from deep soil layers may increase resource use efficiency while maintaining yield under stressed conditions. Winter oilseed rape ( Brassica napus L.) can develop roots access deep-stored resources such as N to sustain its growth productivity. Less is known of the performance varying availability. In this study, we aimed evaluate effects reduced supply on for rape. Oilseed plants grown in outdoor rhizotrons were supplied with 240 80 kg ha −1 ,...
ABSTRACT Due to advances in imaging technologies, the rate of marine video and image data collection is drastically increasing. Often these datasets are not analysed their full potential as extracting information for multiple species, such presence surface area, incredibly time-consuming. This study demonstrates a new open-source interactive machine learning tool, RootPainter, analyse large quickly accurately. The tool was initially developed measure plant roots, but here tested on its...
Abstract The scale of root quantification in research is often limited by the time required for sampling, measurement and processing samples. Recent developments Convolutional Neural Networks (CNN) have made faster more accurate plant image analysis possible which may significantly reduce measurement, but challenges remain making these methods accessible to researchers without an in-depth knowledge Machine Learning. We analyzed images acquired from three destructive samplings using...