Explore the vast range of titles available.

The recent hype about artificial intelligence has sparked renewed interest in applying the successful deep learning (DL) methods for image recognition, speech recognition, robotics, strategic games and other application areas to the field of meteorology. There is some evidence that better weather forecasts can be produced by introducing big data mining and neural networks into the weather prediction workflow. Here, we discuss the question of whether it is possible to completely replace the current numerical weather models and data assimilation systems with DL approaches. This discussion entails a review of state-of-the-art machine learning concepts and their applicability to weather data with its pertinent statistical properties. We think that it is not inconceivable that numerical weather models may one day become obsolete, but a number of fundamental breakthroughs are needed before this goal comes into reach. This article is part of the theme issue ‘Machine learning for weather and climate modelling’.

The prediction of precipitation patterns up to 2 h ahead, also known as precipitation nowcasting, at high spatiotemporal resolutions is of great relevance in weather-dependent decision-making and early warning systems. In this study, we are aiming to provide an efficient and easy-to-understand deep neural network – CLGAN (convolutional long short-term memory generative adversarial network) – to improve the nowcasting skills of heavy precipitation events. The model constitutes a generative adversarial network (GAN) architecture, whose generator is built upon a u-shaped encoder–decoder network (U-Net) and is equipped with recurrent long short-term memory (LSTM) cells to capture spatiotemporal features. The optical flow model DenseRotation and the competitive video prediction models ConvLSTM (convolutional LSTM) and PredRNN-v2 (predictive recurrent neural network version 2) are used as the competitors. A series of evaluation metrics, including the root mean square error, the critical success index, the fractions skill score, and object-based diagnostic evaluation, are utilized for a comprehensive comparison against competing baseline models. We show that CLGAN outperforms the competitors in terms of scores for dichotomous events and object-based diagnostics. A sensitivity analysis on the weight of the GAN component indicates that the GAN-based architecture helps to capture heavy precipitation events. The results encourage future work based on the proposed CLGAN architecture to improve the precipitation nowcasting and early warning systems.

Numerical weather prediction (NWP) models are atmospheric simulations that imitate the dynamics of the atmosphere and provide high-quality forecasts. One of the most significant limitations of NWP is the elevated amount of computational resources required for its functioning, which limits the spatial and temporal resolution of the outputs. Traditional meteorological techniques to increase the resolution are uniquely based on information from a limited group of interest variables. In this study, we offer an alternative approach to the task where we generate precipitation maps based on the complete set of variables of the NWP to generate high-resolution and short-time precipitation predictions. To achieve this, five different deep learning models were trained and evaluated: a baseline, U-Net, two deconvolution networks and one conditional generative model (Conditional Generative Adversarial Network; CGAN). A total of 20 independent random initializations were performed for each of the models. The predictions were evaluated using skill scores based on mean absolute error (MAE) and linear error in probability space (LEPS), equitable threat score (ETS), critical success index (CSI) and frequency bias after applying several thresholds. The models showed a significant improvement in predicting precipitation, showing the benefits of including the complete information from the NWP. The algorithms doubled the resolution of the predictions and corrected an over-forecast bias from the input information. However, some new models presented new types of bias: U-Net tended to mid-range precipitation events, and the deconvolution models favored low rain events and generated some spatial smoothing. The CGAN offered the highest-quality precipitation forecast, generating realistic outputs and indicating possible future research paths.

Purpose: To present experience with cataract extraction in 9 eyes of 7 pediatric patients with chronic uveitis and compare the technique of anterior optic capture in 5 eyes that underwent cataract extraction without optic capture of the intraocular lens (IOL) or were left aphakic. Methods: A retrospective review of pediatric patients with chronic uveitis undergoing cataract surgery was performed, examining the preoperative and postoperative visual acuity, immunosuppressive therapy, surgical technique, complications, subsequent procedures, and need for escalation of systemic immunosuppressive therapy. The technique of anterior optic capture is described in detail. Results: Of the 9 eyes, 5 underwent cataract extraction with IOL placement with the haptics in the capsular bag and optic prolapsed through the anterior capsulorhexis. One eye underwent cataract extraction with IOL implantation in the bag. Three eyes had lensectomy without IOL placement. The eyes with anterior optic capture had no adverse outcomes and uveitis flares were controlled with topical medications and systemic immunosuppressants; the eye with IOL placement without optic capture had recurrent membranes and uveitis flares, necessitating increased systemic immunosuppression. All eyes achieved best-corrected visual acuity of 20/60 or better by 6 months following surgery and 20/30 or better at the most recent follow-up. Conclusions: The technique of cataract extraction with IOL placement and anterior prolapse of the optic through the anterior capsulorhexis shows promise to be a safe and viable option for pediatric patients with chronic uveitis treated with systemic immunotherapy. 2015;52(2):119–125.] [J Pediatr Ophthalmol Strabismus. 2015;52(2):119–125.]

Numerical weather prediction (NWP) models solve a system of partial differential equations based on physical laws to forecast the future state of the atmosphere. These models are deployed operationally, but they are computationally very expensive. Recently, the potential of deep neural networks to generate bespoke weather forecasts has been explored in a couple of scientific studies inspired by the success of video frame prediction models in computer vision. In this study, a simple recurrent neural network with convolutional filters, called ConvLSTM, and an advanced generative network, the Stochastic Adversarial Video Prediction (SAVP) model, are applied to create hourly forecasts of the 2 m temperature for the next 12 h over Europe. We make use of 13 years of data from the ERA5 reanalysis, of which 11 years are utilized for training and 1 year each is used for validating and testing. We choose the 2 m temperature, total cloud cover, and the 850 hPa temperature as predictors and show that both models attain predictive skill by outperforming persistence forecasts. SAVP is superior to ConvLSTM in terms of several evaluation metrics, confirming previous results from computer vision that larger, more complex networks are better suited to learn complex features and to generate better predictions. The 12 h forecasts of SAVP attain a mean squared error (MSE) of about 2.3 K2, an anomaly correlation coefficient (ACC) larger than 0.85, a structural similarity index (SSIM) of around 0.72, and a gradient ratio (rG) of about 0.82. The ConvLSTM yields a higher MSE (3.6 K2), a smaller ACC (0.80) and SSIM (0.65), and a slightly larger rG (0.84). The superior performance of SAVP in terms of MSE, ACC, and SSIM can be largely attributed to the generator. A sensitivity study shows that a larger weight of the generative adversarial network (GAN) component in the SAVP loss leads to even better preservation of spatial variability at the cost of a somewhat increased MSE (2.5 K2). Including the 850 hPa temperature as an additional predictor enhances the forecast quality, and the model also benefits from a larger spatial domain. By contrast, adding the total cloud cover as predictor or reducing the amount of training data to 8 years has only small effects. Although the temperature forecasts obtained in this way are still less powerful than contemporary NWP models, this study demonstrates that sophisticated deep neural networks may achieve considerable forecast quality beyond the nowcasting range in a purely data-driven way.

Purpose: In this study, the authors examined changes in tongue motion caused by glossectomy surgery. A speech task that involved subtle changes in tongue-tip positioning (the motion from /i/ to /s/) was measured. The hypothesis was that patients would have limited motion on the tumor (resected) side and would compensate with greater motion on the nontumor side in order to elevate the tongue tip and blade for /s/. Method: Velocity fields were extracted from tagged magnetic resonance images in the left, middle, and right tongue of 3 patients and 10 controls. Principal components (PCs) analysis quantified motion differences and distinguished between the subject groups. Results: PCs 1 and 2 represented variance in (a) size and independence of the tongue tip, and (b) direction of motion of the tip, body, or both. Patients and controls were correctly separated by a small number of PCs. Conclusions: Motion of the tumor slice was different between patients and controls, but the nontumor side of the patients' tongues did not show excessive or adaptive motion. Both groups contained apical and laminal /s/ users, and 1 patient created apical /s/ in a highly unusual manner.

Aim: To ascertain whether specific testing for “isolated” anti-52 kDa SSA/Ro antibodies (a-SSA/Ro52) during standard anti-extractable nuclear antigen (ENA) testing is clinically useful. Methods: 1438 consecutive sera submitted for anti-ENA testing over 1 year were evaluated for a-SSA/Ro52 using various assays. Results: 7 of 1438 (0.48%) patients were found to have a-SSA/Ro52 without SSA/Ro60 antibodies. Subsequent testing detected a further five patients. Clinical follow-up was possible in 10/12 patients. 2 of these 10 patients had evidence of primary Sjögren’s syndrome (SS) and one had systemic lupus erythematosus (SLE), with sicca symptoms and abnormal Schirmer’s tests. Five other patients had sicca symptoms, of which four had abnormal Schirmer’s tests. Conclusions: “Isolated” anti-52 kDa SSA/Ro antibodies were detected in approximately 0.5% of standard anti-ENA requests, in which their presence was generally not associated with underlying SS or SLE. In view of the increased testing complexity and costs in detecting and confirming these antibodies, specific testing for isolated a-SSA Ro52 antibodies during standard anti-ENA testing seems to be of limited clinical value in a non-obstetric population.

The recent hype about artificial intelligence has sparked renewed interest in applying the successful deep learning (DL) methods for image recognition, speech recognition, robotics, strategic games and other application areas to the field of meteorology. There is some evidence that better weather forecasts can be produced by introducing big data mining and neural networks into the weather prediction workflow. Here, we discuss the question of whether it is possible to completely replace the current numerical weather models and data assimilation systems with DL approaches. This discussion entails a review of state-of-the-art machine learning concepts and their applicability to weather data with its pertinent statistical properties. We think that it is not inconceivable that numerical weather models may one day become obsolete, but a number of fundamental breakthroughs are needed before this goal comes into reach. This article is part of the theme issue ‘Machine learning for weather and climate modelling’.

MRI subsequently showed extensive haemorrhagic change in the muscles of the right thigh (fig 1), and computed tomography of the abdomen showed haemorrhage in the rectus sheath and oblique muscles (fig 2). Significant muscle haemorrhage has not been previously reported in patients with myositis, though recently has been found in a patient receiving therapeutic low molecular weight heparin (LMWH) and warfarin. 1 Heparin treatment is known to be associated with an increased risk of major bleeding, about 0.3%, 2 with both UFH and LMWH, mostly occurring in the gastrointestinal tract, though haemorrhage at the site of injection has been reported. 3 Given the serious nature of this event, we would advise caution in using prophylactic heparin in patients with acute myositis.

Tinnitus is the perception of sound in the absence of a corresponding external acoustic stimulus. With prevalence ranging from 10% to 15%, tinnitus is a common disorder. Many people habituate to the phantom sound, but tinnitus severely impairs quality of life of about 1–2% of all people. Tinnitus has traditionally been regarded as an otological disorder, but advances in neuroimaging methods and development of animal models have increasingly shifted the perspective towards its neuronal correlates. Increased neuronal firing rate, enhanced neuronal synchrony, and changes in the tonotopic organisation are recorded in central auditory pathways in reaction to deprived auditory input and represent—together with changes in non-auditory brain areas—the neuronal correlate of tinnitus. Assessment of patients includes a detailed case history, measurement of hearing function, quantification of tinnitus severity, and identification of causal factors, associated symptoms, and comorbidities. Most widely used treatments for tinnitus involve counselling, and best evidence is available for cognitive behavioural therapy. New pathophysiological insights have prompted the development of innovative brain-based treatment approaches to directly target the neuronal correlates of tinnitus.