Explore the vast range of titles available.

Immunohistochemistry (IHC) is a diagnostic technique used throughout pathology. A machine learning algorithm that could predict individual cell immunophenotype based on hematoxylin and eosin (H&E) staining would save money, time, and reduce tissue consumed. Prior approaches have lacked the spatial accuracy needed for cell-specific analytical tasks. Here IHC performed on destained H&E slides is used to create a neural network that is potentially capable of predicting individual cell immunophenotype. Twelve slides were stained with H&E and scanned to create digital whole slide images. The H&E slides were then destained, and stained with SOX10 IHC. The SOX10 IHC slides were scanned, and corresponding H&E and IHC digital images were registered. Color-thresholding and machine learning techniques were applied to the registered H&E and IHC images to segment 3,396,668 SOX10-negative cells and 306,166 SOX10-positive cells. The resulting segmentation was used to annotate the original H&E images, and a convolutional neural network was trained to predict SOX10 nuclear staining. Sixteen thousand three hundred and nine image patches were used to train the virtual IHC (vIHC) neural network, and 1,813 image patches were used to quantitatively evaluate it. The resulting vIHC neural network achieved an area under the curve of 0.9422 in a receiver operator characteristics analysis when sorting individual nuclei. The vIHC network was applied to additional images from clinical practice, and was evaluated qualitatively by a board-certified dermatopathologist. Further work is needed to make the process more efficient and accurate for clinical use. This proof-of-concept demonstrates the feasibility of creating neural network-driven vIHC assays.

Internal oceanic waves are subsurface gravity waves that can be enormous and travel thousands of kilometres before breaking but they are difficult to study; here observations of such waves in the South China Sea reveal their formation mechanism, extreme turbulence, relationship to the Kuroshio Current and energy budget. IWISE catches internal waves mid-ocean Internal waves are the underwater version of more familiar surface waves. They can be enormous and travel thousands of kilometres before breaking. The South China Sea is known to be home to the largest internal waves in the world's oceans, but their size, generation mechanisms and role in the regional energy budget are unknown. Matthew Alford and colleagues now present the results from the IWISE observational campaign and reveal that internal waves more than 200 metres high break in the South China Sea and create turbulence that is orders of magnitude larger than in the open ocean, and that wave formation is influenced by the Kuroshio current. These results now allow for a complete energy budget of the South China Sea, and for a more accurate incorporation of internal waves into climate models. Internal gravity waves, the subsurface analogue of the familiar surface gravity waves that break on beaches, are ubiquitous in the ocean. Because of their strong vertical and horizontal currents, and the turbulent mixing caused by their breaking, they affect a panoply of ocean processes, such as the supply of nutrients for photosynthesis 1 , sediment and pollutant transport 2 and acoustic transmission 3 ; they also pose hazards for man-made structures in the ocean 4 . Generated primarily by the wind and the tides, internal waves can travel thousands of kilometres from their sources before breaking 5 , making it challenging to observe them and to include them in numerical climate models, which are sensitive to their effects 6 , 7 . For over a decade, studies 8 , 9 , 10 , 11 have targeted the South China Sea, where the oceans’ most powerful known internal waves are generated in the Luzon Strait and steepen dramatically as they propagate west. Confusion has persisted regarding their mechanism of generation, variability and energy budget, however, owing to the lack of in situ data from the Luzon Strait, where extreme flow conditions make measurements difficult. Here we use new observations and numerical models to (1) show that the waves begin as sinusoidal disturbances rather than arising from sharp hydraulic phenomena, (2) reveal the existence of >200-metre-high breaking internal waves in the region of generation that give rise to turbulence levels >10,000 times that in the open ocean, (3) determine that the Kuroshio western boundary current noticeably refracts the internal wave field emanating from the Luzon Strait, and (4) demonstrate a factor-of-two agreement between modelled and observed energy fluxes, which allows us to produce an observationally supported energy budget of the region. Together, these findings give a cradle-to-grave picture of internal waves on a basin scale, which will support further improvements of their representation in numerical climate predictions.

A wide variety of oceanic and atmospheric phenomena are often observed in and around the sunglint region on optical images of the sea surface. The appearance of these phenomena depends strongly on the viewing geometry with areas on the sea surface that are rougher (or smoother) than the background appearing as either brighter or darker than the background depending on their position relative to the specular point. To understand these sea surface signature variations, this paper introduces the concept of a critical sensor viewing angle, defined as the sensor zenith angle at which different sea surface roughness variances produce identical sunglint radiance. It is when the imaging geometry transitions through the critical angle that a surface feature goes through a brightness reversal. Knowledge of where this transition takes place is important for properly interpreting the characteristics of the sea surface signature of these phenomena. The theory behind the concept of the critical angle is presented and then applied to sunglint imagery acquired over the ocean from space by the Moderate Resolution Imaging Spectroradiometer onboard NASA's Aqua and Terra satellites.

Context: Celiac disease (CD) prevalence and diagnosis have increased substantially in recent years. The current gold standard for CD confirmation is visual examination of duodenal mucosal biopsies. An accurate computer-aided biopsy analysis system using deep learning can help pathologists diagnose CD more efficiently. Subjects and Methods: In this study, we trained a deep learning model to detect CD on duodenal biopsy images. Our model uses a state-of-the-art residual convolutional neural network to evaluate patches of duodenal tissue and then aggregates those predictions for whole-slide classification. We tested the model on an independent set of 212 images and evaluated its classification results against reference standards established by pathologists. Results: Our model identified CD, normal tissue, and nonspecific duodenitis with accuracies of 95.3%, 91.0%, and 89.2%, respectively. The area under the receiver operating characteristic curve was >0.95 for all classes. Conclusions: We have developed an automated biopsy analysis system that achieves high performance in detecting CD on biopsy slides. Our system can highlight areas of interest and provide preliminary classification of duodenal biopsies before review by pathologists. This technology has great potential for improving the accuracy and efficiency of CD diagnosis.

Climate change is expected to modify rainfall, temperature and catchment hydrological responses across the world, and adapting to these water-related changes is a pressing challenge. This paper reviews the impact of anthropogenic climate change on water in the UK and looks at projections of future change. The natural variability of the UK climate makes change hard to detect; only historical increases in air temperature can be attributed to anthropogenic climate forcing, but over the last 50 years more winter rainfall has been falling in intense events. Future changes in rainfall and evapotranspiration could lead to changed flow regimes and impacts on water quality, aquatic ecosystems and water availability. Summer flows may decrease on average, but floods may become larger and more frequent. River and lake water quality may decline as a result of higher water temperatures, lower river flows and increased algal blooms in summer, and because of higher flows in the winter. In communicating this important work, researchers should pay particular attention to explaining confidence and uncertainty clearly. Much of the relevant research is either global or highly localized: decision-makers would benefit from more studies that address water and climate change at a spatial and temporal scale appropriate for the decisions they make.

Indian river basins are intensively managed with country-specific agricultural practices of cultivating submerged paddy and uncontrolled groundwater (GW) irrigation. Numerical experiments with the state-of-the-art land surface models, such as variable infiltration capacity (VIC), without incorporating region-specific practices, could be misleading. Here, we coupled VIC with 2D GW model AMBHAS, incorporating India-specific irrigation practices and crop practices, including submerged paddy fields. We performed numerical experiments to understand the causal factors of GW depletion in the northwest Indo-Gangetic plain (IGP). We identify widespread flood irrigation and cultivation of water-intensive paddy as critical drivers of the declining GW scenario. Our numerical experiments suggest that the introduction of drip irrigation reduces GW depletion in the northwest, but does not change the sign of GW level trends. The GW levels in the non-paddy fields of the middle IGP are less sensitive to irrigation practices due to the high return flow to GW for flood irrigation.

Better representations of groundwater processes need to be incorporated into large-scale hydrological models to improve simulations of regional- to global-scale hydrology and climate, as well as understanding of feedbacks between the human and natural systems. We incorporated a 2D groundwater flow model into the variable infiltration capacity (VIC) hydrological model code to address its lack of a lateral groundwater flow component. The water table was coupled with the variably saturated VIC soil column allowing bi-directional exchange of water between the aquifer and the soil. We then investigated how variations in aquifer properties and grid resolution affect modelled evapotranspiration (ET), runoff and groundwater recharge. We simulated nine idealised, homogenous aquifers with different combinations of transmissivity, storage coefficient, and three grid resolutions. The magnitude of cell ET, runoff, and recharge significantly depends on water table depth. In turn, the distribution of water table depths varied significantly as grid resolution increased from 1° to 0.05° for the medium and high transmissivity systems, resulting in changes of model-average fluxes of up to 12.3% of mean rainfall. For the low transmissivity aquifer, increasing the grid resolution has a minimal effect as lateral groundwater flow is low, and the VIC grid cells behave as vertical columns. The inclusion of the 2D groundwater model in VIC will enable the future representation of irrigation from groundwater pumping, and the feedbacks between groundwater use and the hydrological cycle.

Histopathology image evaluation is indispensable for cancer diagnoses and subtype classification. Standard artificial intelligence methods for histopathology image analyses have focused on optimizing specialized models for each diagnostic task 1 , 2 . Although such methods have achieved some success, they often have limited generalizability to images generated by different digitization protocols or samples collected from different populations 3 . Here, to address this challenge, we devised the Clinical Histopathology Imaging Evaluation Foundation (CHIEF) model, a general-purpose weakly supervised machine learning framework to extract pathology imaging features for systematic cancer evaluation. CHIEF leverages two complementary pretraining methods to extract diverse pathology representations: unsupervised pretraining for tile-level feature identification and weakly supervised pretraining for whole-slide pattern recognition. We developed CHIEF using 60,530 whole-slide images spanning 19 anatomical sites. Through pretraining on 44 terabytes of high-resolution pathology imaging datasets, CHIEF extracted microscopic representations useful for cancer cell detection, tumour origin identification, molecular profile characterization and prognostic prediction. We successfully validated CHIEF using 19,491 whole-slide images from 32 independent slide sets collected from 24 hospitals and cohorts internationally. Overall, CHIEF outperformed the state-of-the-art deep learning methods by up to 36.1%, showing its ability to address domain shifts observed in samples from diverse populations and processed by different slide preparation methods. CHIEF provides a generalizable foundation for efficient digital pathology evaluation for patients with cancer. A study describes the development of a generalizable foundation machine learning framework to extract pathology imaging features for cancer diagnosis and prognosis prediction.

Poor water quality is a widespread issue in urban rivers and streams in London. Localised pollution can have impacts on local communities, from health issues to environmental degradation and restricted recreational use of water. The Salmons and Pymmes Brooks, located in the London Borough of Enfield, flow into the River Lee, and in this paper, the impacts of misconnected sewers, urban runoff and atmospheric pollution have been evaluated. The first step towards finding a sustainable and effective solution to these issues is to identify sources and paths of pollutants and to understand their cycle through catchments and rivers. The INCA water quality model has been applied to the Salmons and Pymmes urban catchments in north-east London, with the aim of providing local communities and community action groups such as Thames21 with a tool they can use to assess the water quality issue. INCA is a process-based, dynamic flow and quality model, and so it can account for daily changes in temperature, flow, water velocity and residence time that all affect reaction kinetics and hence chemical flux. As INCA is process-based, a set of mitigation strategies have been evaluated including constructed wetland across the catchment to assess pollution control. The constructed wetlands can make a significant difference reducing sediment transport and improving nutrient control for nitrogen and phosphorus. The results of this paper show that a substantial reduction in nitrate, ammonium and phosphorus concentrations can be achieved if a proper catchment-scale wetland implementation strategy is put in place. Furthermore, the paper shows how the nutrient reduction efficiency of the wetlands should not be affected by climate change.