TY - JOUR A1 - Baumhoer, Celia A. A1 - Dietz, Andreas J. A1 - Kneisel, C. A1 - Kuenzer, C. T1 - Automated Extraction of Antarctic Glacier and Ice Shelf Fronts from Sentinel-1 Imagery Using Deep Learning JF - Remote Sensing N2 - Sea level rise contribution from the Antarctic ice sheet is influenced by changes in glacier and ice shelf front position. Still, little is known about seasonal glacier and ice shelf front fluctuations as the manual delineation of calving fronts from remote sensing imagery is very time-consuming. The major challenge of automatic calving front extraction is the low contrast between floating glacier and ice shelf fronts and the surrounding sea ice. Additionally, in previous decades, remote sensing imagery over the often cloud-covered Antarctic coastline was limited. Nowadays, an abundance of Sentinel-1 imagery over the Antarctic coastline exists and could be used for tracking glacier and ice shelf front movement. To exploit the available Sentinel-1 data, we developed a processing chain allowing automatic extraction of the Antarctic coastline from Seninel-1 imagery and the creation of dense time series to assess calving front change. The core of the proposed workflow is a modified version of the deep learning architecture U-Net. This convolutional neural network (CNN) performs a semantic segmentation on dual-pol Sentinel-1 data and the Antarctic TanDEM-X digital elevation model (DEM). The proposed method is tested for four training and test areas along the Antarctic coastline. The automatically extracted fronts deviate on average 78 m in training and 108 m test areas. Spatial and temporal transferability is demonstrated on an automatically extracted 15-month time series along the Getz Ice Shelf. Between May 2017 and July 2018, the fronts along the Getz Ice Shelf show mostly an advancing tendency with the fastest moving front of DeVicq Glacier with 726 ± 20 m/yr. KW - Antarctica KW - coastline KW - deep learning KW - semantic segmentation KW - Getz Ice Shelf KW - calving front KW - glacier front KW - U-Net KW - convolutional neural network KW - glacier terminus Y1 - 2019 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:20-opus-193150 SN - 2072-4292 VL - 11 IS - 21 ER - TY - JOUR A1 - Pook, Torsten A1 - Freudenthal, Jan A1 - Korte, Arthur A1 - Simianer, Henner T1 - Using Local Convolutional Neural Networks for Genomic Prediction JF - Frontiers in Genetics N2 - The prediction of breeding values and phenotypes is of central importance for both livestock and crop breeding. In this study, we analyze the use of artificial neural networks (ANN) and, in particular, local convolutional neural networks (LCNN) for genomic prediction, as a region-specific filter corresponds much better with our prior genetic knowledge on the genetic architecture of traits than traditional convolutional neural networks. Model performances are evaluated on a simulated maize data panel (n = 10,000; p = 34,595) and real Arabidopsis data (n = 2,039; p = 180,000) for a variety of traits based on their predictive ability. The baseline LCNN, containing one local convolutional layer (kernel size: 10) and two fully connected layers with 64 nodes each, is outperforming commonly proposed ANNs (multi layer perceptrons and convolutional neural networks) for basically all considered traits. For traits with high heritability and large training population as present in the simulated data, LCNN are even outperforming state-of-the-art methods like genomic best linear unbiased prediction (GBLUP), Bayesian models and extended GBLUP, indicated by an increase in predictive ability of up to 24%. However, for small training populations, these state-of-the-art methods outperform all considered ANNs. Nevertheless, the LCNN still outperforms all other considered ANNs by around 10%. Minor improvements to the tested baseline network architecture of the LCNN were obtained by increasing the kernel size and of reducing the stride, whereas the number of subsequent fully connected layers and their node sizes had neglectable impact. Although gains in predictive ability were obtained for large scale data sets by using LCNNs, the practical use of ANNs comes with additional problems, such as the need of genotyping all considered individuals, the lack of estimation of heritability and reliability. Furthermore, breeding values are additive by design, whereas ANN-based estimates are not. However, ANNs also comes with new opportunities, as networks can easily be extended to account for additional inputs (omics, weather etc.) and outputs (multi-trait models), and computing time increases linearly with the number of individuals. With advances in high-throughput phenotyping and cheaper genotyping, ANNs can become a valid alternative for genomic prediction. KW - phenotype prediction KW - Keras KW - genomic selection KW - selection KW - breeding KW - machine learning KW - deep learning Y1 - 2020 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:20-opus-216436 VL - 11 ER - TY - JOUR A1 - Hoeser, Thorsten A1 - Bachofer, Felix A1 - Kuenzer, Claudia T1 - Object detection and image segmentation with deep learning on Earth Observation data: a review — part II: applications JF - Remote Sensing N2 - In Earth observation (EO), large-scale land-surface dynamics are traditionally analyzed by investigating aggregated classes. The increase in data with a very high spatial resolution enables investigations on a fine-grained feature level which can help us to better understand the dynamics of land surfaces by taking object dynamics into account. To extract fine-grained features and objects, the most popular deep-learning model for image analysis is commonly used: the convolutional neural network (CNN). In this review, we provide a comprehensive overview of the impact of deep learning on EO applications by reviewing 429 studies on image segmentation and object detection with CNNs. We extensively examine the spatial distribution of study sites, employed sensors, used datasets and CNN architectures, and give a thorough overview of applications in EO which used CNNs. Our main finding is that CNNs are in an advanced transition phase from computer vision to EO. Upon this, we argue that in the near future, investigations which analyze object dynamics with CNNs will have a significant impact on EO research. With a focus on EO applications in this Part II, we complete the methodological review provided in Part I. KW - artificial intelligence KW - AI KW - machine learning KW - deep learning KW - neural networks KW - convolutional neural networks KW - CNN KW - image segmentation KW - object detection KW - earth observation Y1 - 2020 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:20-opus-213152 SN - 2072-4292 VL - 12 IS - 18 ER - TY - JOUR A1 - Hoeser, Thorsten A1 - Kuenzer, Claudia T1 - Object detection and image segmentation with deep learning on Earth observation data: a review-part I: evolution and recent trends JF - Remote Sensing N2 - Deep learning (DL) has great influence on large parts of science and increasingly established itself as an adaptive method for new challenges in the field of Earth observation (EO). Nevertheless, the entry barriers for EO researchers are high due to the dense and rapidly developing field mainly driven by advances in computer vision (CV). To lower the barriers for researchers in EO, this review gives an overview of the evolution of DL with a focus on image segmentation and object detection in convolutional neural networks (CNN). The survey starts in 2012, when a CNN set new standards in image recognition, and lasts until late 2019. Thereby, we highlight the connections between the most important CNN architectures and cornerstones coming from CV in order to alleviate the evaluation of modern DL models. Furthermore, we briefly outline the evolution of the most popular DL frameworks and provide a summary of datasets in EO. By discussing well performing DL architectures on these datasets as well as reflecting on advances made in CV and their impact on future research in EO, we narrow the gap between the reviewed, theoretical concepts from CV and practical application in EO. KW - artificial intelligence KW - AI KW - machine learning KW - deep learning KW - neural networks KW - convolutional neural networks KW - CNN KW - image segmentation KW - object detection KW - Earth observation Y1 - 2020 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:20-opus-205918 SN - 2072-4292 VL - 12 IS - 10 ER - TY - JOUR A1 - Ankenbrand, Markus Johannes A1 - Lohr, David A1 - Schlötelburg, Wiebke A1 - Reiter, Theresa A1 - Wech, Tobias A1 - Schreiber, Laura Maria T1 - Deep learning-based cardiac cine segmentation: Transfer learning application to 7T ultrahigh-field MRI JF - Magnetic Resonance in Medicine N2 - Purpose Artificial neural networks show promising performance in automatic segmentation of cardiac MRI. However, training requires large amounts of annotated data and generalization to different vendors, field strengths, sequence parameters, and pathologies is limited. Transfer learning addresses this challenge, but specific recommendations regarding type and amount of data required is lacking. In this study, we assess data requirements for transfer learning to experimental cardiac MRI at 7T where the segmentation task can be challenging. In addition, we provide guidelines, tools, and annotated data to enable transfer learning approaches by other researchers and clinicians. Methods A publicly available segmentation model was used to annotate a publicly available data set. This labeled data set was subsequently used to train a neural network for segmentation of left ventricle and myocardium in cardiac cine MRI. The network is used as starting point for transfer learning to 7T cine data of healthy volunteers (n = 22; 7873 images) by updating the pre-trained weights. Structured and random data subsets of different sizes were used to systematically assess data requirements for successful transfer learning. Results Inconsistencies in the publically available data set were corrected, labels created, and a neural network trained. On 7T cardiac cine images the model pre-trained on public imaging data, acquired at 1.5T and 3T, achieved DICE\(_{LV}\) = 0.835 and DICE\(_{MY}\) = 0.670. Transfer learning using 7T cine data and ImageNet weight initialization improved model performance to DICE\(_{LV}\) = 0.900 and DICE\(_{MY}\) = 0.791. Using only end-systolic and end-diastolic images reduced training data by 90%, with no negative impact on segmentation performance (DICE\(_{LV}\) = 0.908, DICE\(_{MY}\) = 0.805). Conclusions This work demonstrates and quantifies the benefits of transfer learning for cardiac cine image segmentation. We provide practical guidelines for researchers planning transfer learning projects in cardiac MRI and make data, models, and code publicly available. KW - 7T KW - ultrahigh-field KW - transfer learning KW - segmentation KW - neural networks KW - deep learning KW - cardiac magnetic resonance KW - cardiac function Y1 - 2021 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:20-opus-257604 VL - 86 IS - 4 ER - TY - JOUR A1 - Dirscherl, Mariel A1 - Dietz, Andreas J. A1 - Kneisel, Christof A1 - Kuenzer, Claudia T1 - A novel method for automated supraglacial lake mapping in Antarctica using Sentinel-1 SAR imagery and deep learning JF - Remote Sensing N2 - Supraglacial meltwater accumulation on ice sheets can be a main driver for accelerated ice discharge, mass loss, and global sea-level-rise. With further increasing surface air temperatures, meltwater-induced hydrofracturing, basal sliding, or surface thinning will cumulate and most likely trigger unprecedented ice mass loss on the Greenland and Antarctic ice sheets. While the Greenland surface hydrological network as well as its impacts on ice dynamics and mass balance has been studied in much detail, Antarctic supraglacial lakes remain understudied with a circum-Antarctic record of their spatio-temporal development entirely lacking. This study provides the first automated supraglacial lake extent mapping method using Sentinel-1 synthetic aperture radar (SAR) imagery over Antarctica and complements the developed optical Sentinel-2 supraglacial lake detection algorithm presented in our companion paper. In detail, we propose the use of a modified U-Net for semantic segmentation of supraglacial lakes in single-polarized Sentinel-1 imagery. The convolutional neural network (CNN) is implemented with residual connections for optimized performance as well as an Atrous Spatial Pyramid Pooling (ASPP) module for multiscale feature extraction. The algorithm is trained on 21,200 Sentinel-1 image patches and evaluated in ten spatially or temporally independent test acquisitions. In addition, George VI Ice Shelf is analyzed for intra-annual lake dynamics throughout austral summer 2019/2020 and a decision-level fused Sentinel-1 and Sentinel-2 maximum lake extent mapping product is presented for January 2020 revealing a more complete supraglacial lake coverage (~770 km\(^2\)) than the individual single-sensor products. Classification results confirm the reliability of the proposed workflow with an average Kappa coefficient of 0.925 and a F\(_1\)-score of 93.0% for the supraglacial water class across all test regions. Furthermore, the algorithm is applied in an additional test region covering supraglacial lakes on the Greenland ice sheet which further highlights the potential for spatio-temporal transferability. Future work involves the integration of more training data as well as intra-annual analyses of supraglacial lake occurrence across the whole continent and with focus on supraglacial lake development throughout a summer melt season and into Antarctic winter. KW - Antarctica KW - Antarctic ice sheet KW - supraglacial lakes KW - ice sheet hydrology KW - Sentinel-1 KW - remote sensing KW - machine learning KW - deep learning KW - semantic segmentation KW - convolutional neural network Y1 - 2021 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:20-opus-222998 SN - 2072-4292 VL - 13 IS - 2 ER - TY - JOUR A1 - Janiesch, Christian A1 - Zschech, Patrick A1 - Heinrich, Kai T1 - Machine learning and deep learning JF - Electronic Markets N2 - Today, intelligent systems that offer artificial intelligence capabilities often rely on machine learning. Machine learning describes the capacity of systems to learn from problem-specific training data to automate the process of analytical model building and solve associated tasks. Deep learning is a machine learning concept based on artificial neural networks. For many applications, deep learning models outperform shallow machine learning models and traditional data analysis approaches. In this article, we summarize the fundamentals of machine learning and deep learning to generate a broader understanding of the methodical underpinning of current intelligent systems. In particular, we provide a conceptual distinction between relevant terms and concepts, explain the process of automated analytical model building through machine learning and deep learning, and discuss the challenges that arise when implementing such intelligent systems in the field of electronic markets and networked business. These naturally go beyond technological aspects and highlight issues in human-machine interaction and artificial intelligence servitization. KW - analytical model building KW - machine learning KW - deep learning KW - artificial intelligence KW - artificial neural networks Y1 - 2021 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:20-opus-270155 SN - 1422-8890 VL - 31 IS - 3 ER - TY - JOUR A1 - Ankenbrand, Markus J. A1 - Shainberg, Liliia A1 - Hock, Michael A1 - Lohr, David A1 - Schreiber, Laura M. T1 - Sensitivity analysis for interpretation of machine learning based segmentation models in cardiac MRI JF - BMC Medical Imaging N2 - Background Image segmentation is a common task in medical imaging e.g., for volumetry analysis in cardiac MRI. Artificial neural networks are used to automate this task with performance similar to manual operators. However, this performance is only achieved in the narrow tasks networks are trained on. Performance drops dramatically when data characteristics differ from the training set properties. Moreover, neural networks are commonly considered black boxes, because it is hard to understand how they make decisions and why they fail. Therefore, it is also hard to predict whether they will generalize and work well with new data. Here we present a generic method for segmentation model interpretation. Sensitivity analysis is an approach where model input is modified in a controlled manner and the effect of these modifications on the model output is evaluated. This method yields insights into the sensitivity of the model to these alterations and therefore to the importance of certain features on segmentation performance. Results We present an open-source Python library (misas), that facilitates the use of sensitivity analysis with arbitrary data and models. We show that this method is a suitable approach to answer practical questions regarding use and functionality of segmentation models. We demonstrate this in two case studies on cardiac magnetic resonance imaging. The first case study explores the suitability of a published network for use on a public dataset the network has not been trained on. The second case study demonstrates how sensitivity analysis can be used to evaluate the robustness of a newly trained model. Conclusions Sensitivity analysis is a useful tool for deep learning developers as well as users such as clinicians. It extends their toolbox, enabling and improving interpretability of segmentation models. Enhancing our understanding of neural networks through sensitivity analysis also assists in decision making. Although demonstrated only on cardiac magnetic resonance images this approach and software are much more broadly applicable. KW - deep learning KW - neural networks KW - cardiac magnetic resonance KW - sensitivity analysis KW - transformations KW - augmentation KW - segmentation Y1 - 2021 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:20-opus-259169 VL - 21 IS - 1 ER - TY - JOUR A1 - Weng, Andreas M. A1 - Heidenreich, Julius F. A1 - Metz, Corona A1 - Veldhoen, Simon A1 - Bley, Thorsten A. A1 - Wech, Tobias T1 - Deep learning-based segmentation of the lung in MR-images acquired by a stack-of-spirals trajectory at ultra-short echo-times JF - BMC Medical Imaging N2 - Background Functional lung MRI techniques are usually associated with time-consuming post-processing, where manual lung segmentation represents the most cumbersome part. The aim of this study was to investigate whether deep learning-based segmentation of lung images which were scanned by a fast UTE sequence exploiting the stack-of-spirals trajectory can provide sufficiently good accuracy for the calculation of functional parameters. Methods In this study, lung images were acquired in 20 patients suffering from cystic fibrosis (CF) and 33 healthy volunteers, by a fast UTE sequence with a stack-of-spirals trajectory and a minimum echo-time of 0.05 ms. A convolutional neural network was then trained for semantic lung segmentation using 17,713 2D coronal slices, each paired with a label obtained from manual segmentation. Subsequently, the network was applied to 4920 independent 2D test images and results were compared to a manual segmentation using the Sørensen–Dice similarity coefficient (DSC) and the Hausdorff distance (HD). Obtained lung volumes and fractional ventilation values calculated from both segmentations were compared using Pearson’s correlation coefficient and Bland Altman analysis. To investigate generalizability to patients outside the CF collective, in particular to those exhibiting larger consolidations inside the lung, the network was additionally applied to UTE images from four patients with pneumonia and one with lung cancer. Results The overall DSC for lung tissue was 0.967 ± 0.076 (mean ± standard deviation) and HD was 4.1 ± 4.4 mm. Lung volumes derived from manual and deep learning based segmentations as well as values for fractional ventilation exhibited a high overall correlation (Pearson’s correlation coefficent = 0.99 and 1.00). For the additional cohort with unseen pathologies / consolidations, mean DSC was 0.930 ± 0.083, HD = 12.9 ± 16.2 mm and the mean difference in lung volume was 0.032 ± 0.048 L. Conclusions Deep learning-based image segmentation in stack-of-spirals based lung MRI allows for accurate estimation of lung volumes and fractional ventilation values and promises to replace the time-consuming step of manual image segmentation in the future. KW - MRI KW - lung KW - deep learning KW - image segmentation Y1 - 2021 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:20-opus-260520 VL - 21 ER - TY - JOUR A1 - Pennig, Lenhard A1 - Hoyer, Ulrike Cornelia Isabel A1 - Krauskopf, Alexandra A1 - Shahzad, Rahil A1 - Jünger, Stephanie T. A1 - Thiele, Frank A1 - Laukamp, Kai Roman A1 - Grunz, Jan-Peter A1 - Perkuhn, Michael A1 - Schlamann, Marc A1 - Kabbasch, Christoph A1 - Borggrefe, Jan A1 - Goertz, Lukas T1 - Deep learning assistance increases the detection sensitivity of radiologists for secondary intracranial aneurysms in subarachnoid hemorrhage JF - Neuroradiology N2 - Purpose To evaluate whether a deep learning model (DLM) could increase the detection sensitivity of radiologists for intracranial aneurysms on CT angiography (CTA) in aneurysmal subarachnoid hemorrhage (aSAH). Methods Three different DLMs were trained on CTA datasets of 68 aSAH patients with 79 aneurysms with their outputs being combined applying ensemble learning (DLM-Ens). The DLM-Ens was evaluated on an independent test set of 104 aSAH patients with 126 aneuryms (mean volume 129.2 ± 185.4 mm3, 13.0% at the posterior circulation), which were determined by two radiologists and one neurosurgeon in consensus using CTA and digital subtraction angiography scans. CTA scans of the test set were then presented to three blinded radiologists (reader 1: 13, reader 2: 4, and reader 3: 3 years of experience in diagnostic neuroradiology), who assessed them individually for aneurysms. Detection sensitivities for aneurysms of the readers with and without the assistance of the DLM were compared. Results In the test set, the detection sensitivity of the DLM-Ens (85.7%) was comparable to the radiologists (reader 1: 91.2%, reader 2: 86.5%, and reader 3: 86.5%; Fleiss κ of 0.502). DLM-assistance significantly increased the detection sensitivity (reader 1: 97.6%, reader 2: 97.6%,and reader 3: 96.0%; overall P=.024; Fleiss κ of 0.878), especially for secondary aneurysms (88.2% of the additional aneurysms provided by the DLM). Conclusion Deep learning significantly improved the detection sensitivity of radiologists for aneurysms in aSAH, especially for secondary aneurysms. It therefore represents a valuable adjunct for physicians to establish an accurate diagnosis in order to optimize patient treatment. KW - aneurysms KW - aneurysmal subarachnoid hemorrhage KW - CT angiography KW - deep learning KW - convolutional neural networks Y1 - 2021 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:20-opus-308117 SN - 0028-3940 SN - 1432-1920 VL - 63 IS - 12 ER -