Alzheimer's and Parkinson's diseases predict different COVID-19 outcomes, a UK Biobank study

In December 2019, a coronavirus, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) began infecting humans causing a novel disease, coronavirus disease 19 (COVID-19). This was first described in the Wuhan province of the Peoples Republic of China. SARS-CoV-2 spread throughout the world causing a global pandemic. To date, thousands of cases of COVID-19 were reported in the United Kingdom, and over 45,000 patients have died. Some progress has been achieved in managing this disease, but the biological determinants of health, besides age, that affect COVID-19 infectivity and mortality are under scrutiny. Recent studies show that several medical conditions, including diabetes and hypertension, increase the risk of COVID-19 infection and death. The increased vulnerability of the elderly and those with comorbidities, together with the prevalence of neurodegenerative diseases with advanced age, led us to investigate the links between neurodegeneration and COVID-19. We analysed the primary health records of 13,338 UK individuals tested for COVID-19 between March and July 2020. We show that a pre-existing diagnosis of Alzheimers disease predicts the highest risk of COVID-19 infection and mortality among the elderly. In contrast, Parkinsons disease patients were found to be at increased risk of infection but not mortality from COVID-19. We conclude that there are disease-specific differences in COVID-19 susceptibility among patients affected by neurodegenerative disorders.

CT-based Rapid Triage of COVID-19 Patients: Risk Prediction and Progression Estimation of ICU Admission, Mechanical Ventilation, and Death of Hospitalized Patients

The wave of COVID-19 continues to overwhelm the medical resources, especially the stressed intensive care unit (ICU) capacity and the shortage of mechanical ventilation (MV). Here we performed CT-based analysis combined with electronic health records and clinical laboratory results on Cohort 1 (n = 1662 from 17 hospitals) with prognostic estimation for the rapid stratification of PCR confirmed COVID-19 patients. These models, validated on Cohort 2 (n = 700) and Cohort 3 (n = 662) constructed from 9 external hospitals, achieved satisfying performance for predicting ICU, MV and death of COVID-19 patients (AUROC 0.916, 0.919 and 0.853), even on events happened two days later after admission (AUROC 0.919, 0.943 and 0.856). Both clinical and image features showed complementary roles in events prediction and provided accurate estimates to the time of progression (p<.001). Our findings are valuable for delivering timely treatment and optimizing the use of medical resources in the pandemic of COVID-19.

Links between air pollution and COVID-19 in England

In December 2019 a novel disease coronavirus disease 19 (COVID-19) emerged in the Wuhan province of the Peoples Republic of China. COVID-19 is caused by a novel coronavirus (SARS-CoV-2) thought to have jumped species, from another mammal to humans. A pandemic caused by this virus is running rampant throughout the world. Thousands of cases of COVID-19 are reported in England and over 10,000 patients have died. Whilst there has been progress in managing this disease, it is not clear which factors, besides age, affect the severity and mortality of COVID-19. A recent analysis of COVID-19 in Italy identified links between air pollution and death rates. Here, we explored the correlation between three major air pollutants linked to fossil fuels and SARS-CoV-2 lethality in England. We compare up-to-date, real-time SARS-CoV-2 cases and death measurements from public databases to air pollution data monitored across over 120 sites in different regions. We found that the levels of some markers of poor air quality, nitrogen oxides and ozone, are associated with COVID-19 lethality in different English regions. We conclude that the levels of some air pollutants are linked to COVID-19 cases and morbidity. We suggest that our study provides a useful framework to guide health policy in countries affected by this pandemic.

Piecewise Flat Embedding for Image Segmentation.

We introduce a new multi-dimensional nonlinear embedding-Piecewise Flat Embedding (PFE)-for image segmentation. Based on the theory of sparse signal recovery, piecewise flat embedding with diverse channels attempts to recover a piecewise constant image representation with sparse region boundaries and sparse cluster value scattering. The resultant piecewise flat embedding exhibits interesting properties such as suppressing slowly varying signals, and offers an image representation with higher region identifiability which is desirable for image segmentation or high-level semantic analysis tasks. We formulate our embedding as a variant of the Laplacian Eigenmap embedding with an $L_{1,p} (0

Advances in application of artificial intelligence in medical image analysis / 中国医学影像技术

Artificial intelligence (AI), represented by deep learning (DL)has made a major breakthrough in computer vision tasks. The applications and developments of AI in medical image analysis were reviewed from four groups corresponding to four classical computer vision tasks, namely, image classification, object detection, semantic segmentation and image synthesis.

Construction and clinical preliminary validation of an automaticbone age assessment model based on deep learning / 中华放射学杂志

Objective@#To build an automatic bone age assessment system based on China 05 Bone Age Standard and the latest deep learning technology, and preliminary clinical verification was carried out.@*Methods@#The left-hand radiographs of 5 000 children with suspected metabolic disorders were acquired from Wuxi Children′s Hospital. Among these cases, 2 351 patients were randomly chosen as training set, and 101 patients were randomly used as validation set. Four professional pediatric radiologists annotated the development stage according to the China 05 RUS-CHN standard with double-blind method. The mean value of the bone age assessed by experts was the reference standard which was used to train and validate the deep learning mothods based artificial intelligence (AI) model. Accuracy, mean absolute error (MAE), root mean squared error (RMSE) and time efficiency of bone age assessment were compared by using Chi-square test and t test and F test between resident doctors and AI model in the validation set.@*Results@#The MAE and RMSE was (0.37±0.35) years and 0.50 years between AI model and reference standard, respeactively. When the error range was within ±1.0, ±0.7 and ±0.5 years, the accuracy of model on the validation set was 94.1% (95/101), 89.1% (90/101), 74.3% (75/101) respectively. The accuracy between two resident doctors and AI prediction wasn′t statistical significant (P>0.05).@*Conclusion@#The AI model of bone age assessment based on deep learning is feasible and has the characteristics of high accuracy and efficiency.

Exemplar-Based Image and Video Stylization Using Fully Convolutional Semantic Features.

Color and tone stylization in images and videos strives to enhance unique themes with artistic color and tone adjustments. It has a broad range of applications from professional image postprocessing to photo sharing over social networks. Mainstream photo enhancement softwares, such as Adobe Lightroom and Instagram, provide users with predefined styles, which are often hand-crafted through a trial-and-error process. Such photo adjustment tools lack a semantic understanding of image contents and the resulting global color transform limits the range of artistic styles it can represent. On the other hand, stylistic enhancement needs to apply distinct adjustments to various semantic regions. Such an ability enables a broader range of visual styles. In this paper, we first propose a novel deep learning architecture for exemplar-based image stylization, which learns local enhancement styles from image pairs. Our deep learning architecture consists of fully convolutional networks for automatic semantics-aware feature extraction and fully connected neural layers for adjustment prediction. Image stylization can be efficiently accomplished with a single forward pass through our deep network. To extend our deep network from image stylization to video stylization, we exploit temporal superpixels to facilitate the transfer of artistic styles from image exemplars to videos. Experiments on a number of data sets for image stylization as well as a diverse set of video clips demonstrate the effectiveness of our deep learning architecture.

ColorSketch: A Drawing Assistant for Generating Color Sketches from Photos.

A color sketch creates a vivid depiction of a scene using sparse pencil strokes and casual colored brush strokes. The interactive drawing system ColorSketch can help novice users generate color sketches from photos. To preserve artistic freedom and expressiveness, the proposed system gives users full control over pencil strokes, while automatically augmenting pencil sketches using color mapping, brush stroke rendering, and blank area creation. Experimental and user study results demonstrate that users, especially novices, can create better color sketches with our system than when using traditional manual tools.

Visual Saliency Detection Based on Multiscale Deep CNN Features.

Visual saliency is a fundamental problem in both cognitive and computational sciences, including computer vision. In this paper, we discover that a high-quality visual saliency model can be learned from multiscale features extracted using deep convolutional neural networks (CNNs), which have had many successes in visual recognition tasks. For learning such saliency models, we introduce a neural network architecture, which has fully connected layers on top of CNNs responsible for feature extraction at three different scales. The penultimate layer of our neural network has been confirmed to be a discriminative high-level feature vector for saliency detection, which we call deep contrast feature. To generate a more robust feature, we integrate handcrafted low-level features with our deep contrast feature. To promote further research and evaluation of visual saliency models, we also construct a new large database of 4447 challenging images and their pixelwise saliency annotations. Experimental results demonstrate that our proposed method is capable of achieving the state-of-the-art performance on all public benchmarks, improving the F-measure by 6.12% and 10%, respectively, on the DUT-OMRON data set and our new data set (HKU-IS), and lowering the mean absolute error by 9% and 35.3%, respectively, on these two data sets.