Search

The tilt of solar active regions described by Joy's law is essential for converting a toroidal field to a poloidal field in Babcock-Leighton dynamo models. In thin flux tube models the Coriolis force causes Joy's law, acting on east-west flows as they rise towards the surface. Our goal is to measure the evolution of the average tilt angle of hundreds of active regions as they emerge, so that we can constrain the origins of Joy's law. We measured the tilt angle of the primary bipoles in 153 emerging active regions in the Solar Dynamics Observatory Helioseismic Emerging Active Region survey. We used line-of-sight magnetic field measurements averaged over 6 hours to define the polarities and measure the tilt angle up to four days after emergence. We find that at the time of emergence the polarities are on average aligned east-west, and that neither the separation nor the tilt depends on latitude. We do find, however, that ARs at higher latitudes have a faster north-south separation speed than those closer to the equator at the emergence time. After emergence, the tilt angle increases and Joy's law is evident about two days later. The scatter in the tilt angle is independent of flux until about one day after emergence, when higher-flux regions have a smaller scatter in tilt angle than lower-flux regions. Our finding that active regions emerge with an east-west alignment is surprising since thin flux tube models predict that tilt angles of rising flux tubes are generated below the surface. Previously reported tilt angle relaxation of deeply anchored flux tubes can be largely explained by the change in east-west separation. We conclude that Joy's law is caused by an inherent north-south separation speed present when the flux first reaches the surface, and that the scatter in the tilt angle is consistent with buffeting of the polarities by supergranulation., Comment: accepted in Astronomy and Astrophysics

Recording muscle tendon junction displacements during movement, allows separate investigation of the muscle and tendon behaviour, respectively. In order to provide a fully-automatic tracking method, we employ a novel deep learning approach to detect the position of the muscle tendon junction in ultrasound images. We utilize the attention mechanism to enable the network to focus on relevant regions and to obtain a better interpretation of the results. Our data set consists of a large cohort of 79 healthy subjects and 28 subjects with movement limitations performing passive full range of motion and maximum contraction movements. Our trained network shows robust detection of the muscle tendon junction on a diverse data set of varying quality with a mean absolute error of 2.55$\pm$1 mm. We show that our approach can be applied for various subjects and can be operated in real-time. The complete software package is available for open-source use via: https://github.com/luuleitner/deepMTJ, Comment: Accepted version to be published in 2020, 42nd Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), Montreal, Canada

Tissue characterisation with CMR parametric mapping has the potential to detect and quantify both focal and diffuse alterations in myocardial structure not assessable by late gadolinium enhancement. Native T1 mapping in particular has shown promise as a useful biomarker to support diagnostic, therapeutic and prognostic decision-making in ischaemic and non-ischaemic cardiomyopathies. Convolutional neural networks with Bayesian inference are a category of artificial neural networks which model the uncertainty of the network output. This study presents an automated framework for tissue characterisation from native ShMOLLI T1 mapping at 1.5T using a Probabilistic Hierarchical Segmentation (PHiSeg) network. In addition, we use the uncertainty information provided by the PHiSeg network in a novel automated quality control (QC) step to identify uncertain T1 values. The PHiSeg network and QC were validated against manual analysis on a cohort of the UK Biobank containing healthy subjects and chronic cardiomyopathy patients. We used the proposed method to obtain reference T1 ranges for the left ventricular myocardium in healthy subjects as well as common clinical cardiac conditions. T1 values computed from automatic and manual segmentations were highly correlated (r=0.97). Bland-Altman analysis showed good agreement between the automated and manual measurements. The average Dice metric was 0.84 for the left ventricular myocardium. The sensitivity of detection of erroneous outputs was 91%. Finally, T1 values were automatically derived from 14,683 CMR exams from the UK Biobank. The proposed pipeline allows for automatic analysis of myocardial native T1 mapping and includes a QC process to detect potentially erroneous results. T1 reference values were presented for healthy subjects and common clinical cardiac conditions from the largest cohort to date using T1-mapping images.

In a systematic review, we investigate current applications of ultrasound in locomotion research. Shortcomings in the range of view of ultrasound systems affect the direct validation of musculoskeletal simulations as inverse approaches have to be applied. We present currently used methods to estimate muscle and tendon length in human plantarflexors., Comment: Accepted for publication to CAMS-Knee OpenSim 2020, (4 pages, 2 figures, 1 table)

One of the key drawbacks of 3D convolutional neural networks for segmentation is their memory footprint, which necessitates compromises in the network architecture in order to fit into a given memory budget. Motivated by the RevNet for image classification, we propose a partially reversible U-Net architecture that reduces memory consumption substantially. The reversible architecture allows us to exactly recover each layer's outputs from the subsequent layer's ones, eliminating the need to store activations for backpropagation. This alleviates the biggest memory bottleneck and enables very deep (theoretically infinitely deep) 3D architectures. On the BraTS challenge dataset, we demonstrate substantial memory savings. We further show that the freed memory can be used for processing the whole field-of-view (FOV) instead of patches. Increasing network depth led to higher segmentation accuracy while growing the memory footprint only by a very small fraction, thanks to the partially reversible architecture., Comment: Accepted to MICCAI 2019; Edit v2: Added reference to related work of Blumberg et al

Segmentation of anatomical structures and pathologies is inherently ambiguous. For instance, structure borders may not be clearly visible or different experts may have different styles of annotating. The majority of current state-of-the-art methods do not account for such ambiguities but rather learn a single mapping from image to segmentation. In this work, we propose a novel method to model the conditional probability distribution of the segmentations given an input image. We derive a hierarchical probabilistic model, in which separate latent variables are responsible for modelling the segmentation at different resolutions. Inference in this model can be efficiently performed using the variational autoencoder framework. We show that our proposed method can be used to generate significantly more realistic and diverse segmentation samples compared to recent related work, both, when trained with annotations from a single or multiple annotators., Comment: Accepted to MICCAI 2019

Supervised deep learning methods for segmentation require large amounts of labelled training data, without which they are prone to overfitting, not generalizing well to unseen images. In practice, obtaining a large number of annotations from clinical experts is expensive and time-consuming. One way to address scarcity of annotated examples is data augmentation using random spatial and intensity transformations. Recently, it has been proposed to use generative models to synthesize realistic training examples, complementing the random augmentation. So far, these methods have yielded limited gains over the random augmentation. However, there is potential to improve the approach by (i) explicitly modeling deformation fields (non-affine spatial transformation) and intensity transformations and (ii) leveraging unlabelled data during the generative process. With this motivation, we propose a novel task-driven data augmentation method where to synthesize new training examples, a generative network explicitly models and applies deformation fields and additive intensity masks on existing labelled data, modeling shape and intensity variations, respectively. Crucially, the generative model is optimized to be conducive to the task, in this case segmentation, and constrained to match the distribution of images observed from labelled and unlabelled samples. Furthermore, explicit modeling of deformation fields allow synthesizing segmentation masks and images in exact correspondence by simply applying the generated transformation to an input image and the corresponding annotation. Our experiments on cardiac magnetic resonance images (MRI) showed that, for the task of segmentation in small training data scenarios, the proposed method substantially outperforms conventional augmentation techniques., Comment: 13 pages, 3 figures, 1 table, This article has been accepted at the 26th international conference on Information Processing in Medical Imaging (IPMI) 2019

Accurate segmentation of medical images is an important step towards analyzing and tracking disease related morphological alterations in the anatomy. Convolutional neural networks (CNNs) have recently emerged as a powerful tool for many segmentation tasks in medical imaging. The performance of CNNs strongly depends on the size of the training data and combining data from different sources is an effective strategy for obtaining larger training datasets. However, this is often challenged by heterogeneous labeling of the datasets. For instance, one of the dataset may be missing labels or a number of labels may have been combined into a super label. In this work we propose a cost function which allows integration of multiple datasets with heterogeneous label subsets into a joint training. We evaluated the performance of this strategy on thigh MR and a cardiac MR datasets in which we artificially merged labels for half of the data. We found the proposed cost function substantially outperforms a naive masking approach, obtaining results very close to using the full annotations., Comment: Accepted for presentation at 9th International Conference on Machine Learning in Medical Imaging (MLMI 2018)

Semantic segmentation of medical images is a crucial step for the quantification of healthy anatomy and diseases alike. The majority of the current state-of-the-art segmentation algorithms are based on deep neural networks and rely on large datasets with full pixel-wise annotations. Producing such annotations can often only be done by medical professionals and requires large amounts of valuable time. Training a medical image segmentation network with weak annotations remains a relatively unexplored topic. In this work we investigate training strategies to learn the parameters of a pixel-wise segmentation network from scribble annotations alone. We evaluate the techniques on public cardiac (ACDC) and prostate (NCI-ISBI) segmentation datasets. We find that the networks trained on scribbles suffer from a remarkably small degradation in Dice of only 2.9% (cardiac) and 4.5% (prostate) with respect to a network trained on full annotations., Comment: Accepted for presentation at DLMIA 2018

Convolutional neural networks (CNNs) have shown promising results on several segmentation tasks in magnetic resonance (MR) images. However, the accuracy of CNNs may degrade severely when segmenting images acquired with different scanners and/or protocols as compared to the training data, thus limiting their practical utility. We address this shortcoming in a lifelong multi-domain learning setting by treating images acquired with different scanners or protocols as samples from different, but related domains. Our solution is a single CNN with shared convolutional filters and domain-specific batch normalization layers, which can be tuned to new domains with only a few ($\approx$ 4) labelled images. Importantly, this is achieved while retaining performance on the older domains whose training data may no longer be available. We evaluate the method for brain structure segmentation in MR images. Results demonstrate that the proposed method largely closes the gap to the benchmark, which is training a dedicated CNN for each scanner., Comment: Accepted at the 21st International Conference on Medical Image Computing and Computer Assisted Interventions (MICCAI 2018) in Granada, Spain

Measurement of head biometrics from fetal ultrasonography images is of key importance in monitoring the healthy development of fetuses. However, the accurate measurement of relevant anatomical structures is subject to large inter-observer variability in the clinic. To address this issue, an automated method utilizing Fully Convolutional Networks (FCN) is proposed to determine measurements of fetal head circumference (HC) and biparietal diameter (BPD). An FCN was trained on approximately 2000 2D ultrasound images of the head with annotations provided by 45 different sonographers during routine screening examinations to perform semantic segmentation of the head. An ellipse is fitted to the resulting segmentation contours to mimic the annotation typically produced by a sonographer. The model's performance was compared with inter-observer variability, where two experts manually annotated 100 test images. Mean absolute model-expert error was slightly better than inter-observer error for HC (1.99mm vs 2.16mm), and comparable for BPD (0.61mm vs 0.59mm), as well as Dice coefficient (0.980 vs 0.980). Our results demonstrate that the model performs at a level similar to a human expert, and learns to produce accurate predictions from a large dataset annotated by many sonographers. Additionally, measurements are generated in near real-time at 15fps on a GPU, which could speed up clinical workflow for both skilled and trainee sonographers., Comment: EMBC 2018

We investigated how the magnetic field in solar active regions (ARs) controls flare activity, i.e., whether a confined or eruptive flare occurs. We analyzed 44 flares of GOES class M5.0 and larger that occurred during 2011--2015. We used 3D potential magnetic field models to study their location (using the flare distance from the flux-weighted AR center $d_{\mathrm{FC}}$) and the strength of the magnetic field in the corona above (via decay index $n$ and flux ratio). We also present a first systematic study of the orientation of the coronal magnetic field, using the orientation $\varphi$ of the flare-relevant polarity inversion line as a measure. We analyzed all quantities with respect to the size of the underlying dipole field, characterized by the distance between the opposite-polarity centers, $d_{\mathrm{PC}}$. Flares originating from underneath the AR dipole $(d_{\mathrm{FC}}/d_{\mathrm{PC}}<0.5$) tend to be eruptive if launched from compact ARs ($d_{\mathrm{PC}}\leq60$ Mm) and confined if launched from extended ARs. Flares ejected from the periphery of ARs ($d_{\mathrm{FC}}/d_{\mathrm{PC}}>0.5$) are predominantly eruptive. In confined events the flare-relevant field adjusts its orientation quickly to that of the underlying dipole with height ($\Delta\varphi\gtrsim40^\circ$ until the apex of the dipole field), in contrast to eruptive events where it changes more slowly with height. The critical height for torus instability, $h_{\mathrm{crit}}=h(n=1.5)$, discriminates best between confined ($h_{\mathrm{crit}}\gtrsim40$ Mm) and eruptive flares ($h_{\mathrm{crit}}\lesssim40$ Mm). It discriminates better than $\Delta\varphi$, implying that the decay of the confining field plays a stronger role than its orientation at different heights., Comment: 17 pages, 7 figures, accepted for publication in ApJ

Algorithms for Magnetic Resonance (MR) image reconstruction from undersampled measurements exploit prior information to compensate for missing k-space data. Deep learning (DL) provides a powerful framework for extracting such information from existing image datasets, through learning, and then using it for reconstruction. Leveraging this, recent methods employed DL to learn mappings from undersampled to fully sampled images using paired datasets, including undersampled and corresponding fully sampled images, integrating prior knowledge implicitly. In this article, we propose an alternative approach that learns the probability distribution of fully sampled MR images using unsupervised DL, specifically Variational Autoencoders (VAE), and use this as an explicit prior term in reconstruction, completely decoupling the encoding operation from the prior. The resulting reconstruction algorithm enjoys a powerful image prior to compensate for missing k-space data without requiring paired datasets for training nor being prone to associated sensitivities, such as deviations in undersampling patterns used in training and test time or coil settings. We evaluated the proposed method with T1 weighted images from a publicly available dataset, multi-coil complex images acquired from healthy volunteers (N=8) and images with white matter lesions. The proposed algorithm, using the VAE prior, produced visually high quality reconstructions and achieved low RMSE values, outperforming most of the alternative methods on the same dataset. On multi-coil complex data, the algorithm yielded accurate magnitude and phase reconstruction results. In the experiments on images with white matter lesions, the method faithfully reconstructed the lesions. Keywords: Reconstruction, MRI, prior probability, machine learning, deep learning, unsupervised learning, density estimation, Comment: Published in IEEE TMI. Main text and supplementary material, 19 pages total

Attributing the pixels of an input image to a certain category is an important and well-studied problem in computer vision, with applications ranging from weakly supervised localisation to understanding hidden effects in the data. In recent years, approaches based on interpreting a previously trained neural network classifier have become the de facto state-of-the-art and are commonly used on medical as well as natural image datasets. In this paper, we discuss a limitation of these approaches which may lead to only a subset of the category specific features being detected. To address this problem we develop a novel feature attribution technique based on Wasserstein Generative Adversarial Networks (WGAN), which does not suffer from this limitation. We show that our proposed method performs substantially better than the state-of-the-art for visual attribution on a synthetic dataset and on real 3D neuroimaging data from patients with mild cognitive impairment (MCI) and Alzheimer's disease (AD). For AD patients the method produces compellingly realistic disease effect maps which are very close to the observed effects., Comment: Accepted to CVPR 2018

Accurate segmentation of the heart is an important step towards evaluating cardiac function. In this paper, we present a fully automated framework for segmentation of the left (LV) and right (RV) ventricular cavities and the myocardium (Myo) on short-axis cardiac MR images. We investigate various 2D and 3D convolutional neural network architectures for this task. We investigate the suitability of various state-of-the art 2D and 3D convolutional neural network architectures, as well as slight modifications thereof, for this task. Experiments were performed on the ACDC 2017 challenge training dataset comprising cardiac MR images of 100 patients, where manual reference segmentations were made available for end-diastolic (ED) and end-systolic (ES) frames. We find that processing the images in a slice-by-slice fashion using 2D networks is beneficial due to a relatively large slice thickness. However, the exact network architecture only plays a minor role. We report mean Dice coefficients of $0.950$ (LV), $0.893$ (RV), and $0.899$ (Myo), respectively with an average evaluation time of 1.1 seconds per volume on a modern GPU., Comment: to appear in STACOM 2017 proceedings

We introduce a method of local gating for van der Waals heterostructures, employing a few-layer graphene patterned bottom gate. Being a member of the 2D material family, few-layer graphene adapts perfectly to the commonly used stacking method. Its versatility regarding patterning as well as its flatness make it an ideal candidate for experiments on locally gated 2D materials. Moreover, in combination with ultra-thin hexagonal boron nitride as an insulating layer, sharp potential steps can be created and the quality of the investigated 2D material can be sustained. To underline the good feasibility and performance, we show results on transport experiments in periodically modulated graphene- boron nitride heterostructures, where the charge carrier density is tuned via locally acting patterned few layer graphene bottom gates and a global back gate.

Significant advances have been made towards building accurate automatic segmentation systems for a variety of biomedical applications using machine learning. However, the performance of these systems often degrades when they are applied on new data that differ from the training data, for example, due to variations in imaging protocols. Manually annotating new data for each test domain is not a feasible solution. In this work we investigate unsupervised domain adaptation using adversarial neural networks to train a segmentation method which is more invariant to differences in the input data, and which does not require any annotations on the test domain. Specifically, we learn domain-invariant features by learning to counter an adversarial network, which attempts to classify the domain of the input data by observing the activations of the segmentation network. Furthermore, we propose a multi-connected domain discriminator for improved adversarial training. Our system is evaluated using two MR databases of subjects with traumatic brain injuries, acquired using different scanners and imaging protocols. Using our unsupervised approach, we obtain segmentation accuracies which are close to the upper bound of supervised domain adaptation.

Identifying and interpreting fetal standard scan planes during 2D ultrasound mid-pregnancy examinations are highly complex tasks which require years of training. Apart from guiding the probe to the correct location, it can be equally difficult for a non-expert to identify relevant structures within the image. Automatic image processing can provide tools to help experienced as well as inexperienced operators with these tasks. In this paper, we propose a novel method based on convolutional neural networks which can automatically detect 13 fetal standard views in freehand 2D ultrasound data as well as provide a localisation of the fetal structures via a bounding box. An important contribution is that the network learns to localise the target anatomy using weak supervision based on image-level labels only. The network architecture is designed to operate in real-time while providing optimal output for the localisation task. We present results for real-time annotation, retrospective frame retrieval from saved videos, and localisation on a very large and challenging dataset consisting of images and video recordings of full clinical anomaly screenings. We found that the proposed method achieved an average F1-score of 0.798 in a realistic classification experiment modelling real-time detection, and obtained a 90.09% accuracy for retrospective frame retrieval. Moreover, an accuracy of 77.8% was achieved on the localisation task., Comment: 12 pages, 8 figures, published in IEEE Transactions in Medical Imaging

Multi-atlas segmentation is a widely used tool in medical image analysis, providing robust and accurate results by learning from annotated atlas datasets. However, the availability of fully annotated atlas images for training is limited due to the time required for the labelling task. Segmentation methods requiring only a proportion of each atlas image to be labelled could therefore reduce the workload on expert raters tasked with annotating atlas images. To address this issue, we first re-examine the labelling problem common in many existing approaches and formulate its solution in terms of a Markov Random Field energy minimisation problem on a graph connecting atlases and the target image. This provides a unifying framework for multi-atlas segmentation. We then show how modifications in the graph configuration of the proposed framework enable the use of partially annotated atlas images and investigate different partial annotation strategies. The proposed method was evaluated on two Magnetic Resonance Imaging (MRI) datasets for hippocampal and cardiac segmentation. Experiments were performed aimed at (1) recreating existing segmentation techniques with the proposed framework and (2) demonstrating the potential of employing sparsely annotated atlas data for multi-atlas segmentation.