Deep convolutional neural networks (CNNs) have shown superior performance on the task of single-label image classification. However, the applicability of CNNs to multi-label images still remains an open problem, mainly because of two reasons. First, each image is usually treated as an inseparable entity and represented as one instance, which mixes the visual information corresponding to different labels. Second, the correlations amongst labels are often overlooked. To address these limitations, we propose a deep multi-modal CNN for multi-instance multi-label image classification, called MMCNN-MIML. By combining CNNs with multi-instance multi-label (MIML) learning, our model represents each image as a bag of instances for image classification and inherits the merits of both CNNs and MIML. In particular, MMCNN-MIML has three main appealing properties: 1) it can automatically generate instance representations for MIML by exploiting the architecture of CNNs; 2) it takes advantage of the label correlations by grouping labels in its later layers; and 3) it incorporates the textual context of label groups to generate multi-modal instances, which are effective in discriminating visually similar objects belonging to different groups. Empirical studies on several benchmark multi-label image data sets show that MMCNN-MIML significantly outperforms the state-of-the-art baselines on multi-label image classification tasks.

Background: This study sought to investigate incidence and risk factors for acute kidney injury (AKI) in hospitalized COVID-19. Methods: In this retrospective study, we enrolled 823 COVID-19 patients with at least two evaluations of renal function during hospitalization from four hospitals in Wuhan, China between February 2020 and April 2020. Clinical and laboratory parameters at the time of admission and follow-up data were recorded. Systemic renal tubular dysfunction was evaluated via 24-h urine collections in a subgroup of 55 patients. Results: In total, 823 patients were enrolled (50.5% male) with a mean age of 60.9 ± 14.9 years. AKI occurred in 38 (40.9%) ICU cases but only 6 (0.8%) non-ICU cases. Using forward stepwise Cox regression analysis, we found eight independent risk factors for AKI including decreased platelet level, lower albumin level, lower phosphorus level, higher level of lactate dehydrogenase (LDH), procalcitonin, C-reactive protein (CRP), urea, and prothrombin time (PT) on admission. For every 0.1 mmol/L decreases in serum phosphorus level, patients had a 1.34-fold (95% CI 1.14-1.58) increased risk of AKI. Patients with hypophosphatemia were likely to be older and with lower lymphocyte count, lower serum albumin level, lower uric acid, higher LDH, and higher CRP. Furthermore, serum phosphorus level was positively correlated with phosphate tubular maximum per volume of filtrate (TmP/GFR) (Pearson r ¼ 0.66, p < .001) in subgroup analysis, indicating renal phosphate loss via proximal renal tubular dysfunction. Conclusion:The AKI incidence was very low in non-ICU patients as compared to ICU patients. Hypophosphatemia is an independent risk factor for AKI in patients hospitalized for COVID-19 infection.

The purpose of the General Data Protection Regulation (GDPR) is to provide improved privacy protection. If an app controls personal data from users, it needs to be compliant with GDPR. However, GDPR lists general rules rather than exact step-by-step guidelines about how to develop an app that fulfills the requirements. Therefore, there may exist GDPR compliance violations in existing apps, which would pose severe privacy threats to app users. In this paper, we take mobile health applications (mHealth apps) as a peephole to examine the status quo of GDPR compliance in Android apps. We first propose an automated system, named HPDROID, to bridge the semantic gap between the general rules of GDPR and the app implementations by identifying the data practices declared in the app privacy policy and the data relevant behaviors in the app code. Then, based on HPDROID, we detect three kinds of GDPR compliance violations, including the incompleteness of privacy policy, the inconsistency of data collections, and the insecurity of data transmission. We perform an empirical evaluation of 796 mHealth apps. The results reveal that 189 (23.7%) of them do not provide complete privacy policies. Moreover, 59 apps collect sensitive data through different measures, but 46 (77.9%) of them contain at least one inconsistent collection behavior. Even worse, among the 59 apps, only 8 apps try to ensure the transmission security of collected data. However, all of them contain at least one encryption or SSL misuse. Our work exposes severe privacy issues to raise awareness of privacy protection for app users and developers.

In the era of Big Data, knowledge engineering has to face fundamental challenges by fragmented knowledge from heterogeneous, autonomous sources with complex and evolving relationships. The knowledge representation, knowledge acquisition, and knowledge inference techniques developed in the 1970s and 1980s, driven by research and development of expert systems, need to be updated to cope with both fragmented knowledge from multiple sources in the Big Data revolution, and in-depth expertise from domain experts. This article presents BigKE, a 3phase "online learning -knowledge fusion -knowledge service" knowledge engineering framework with Big Data, with three fundamental research problems: (1) fragmented knowledge modeling and online learning from multiple information sources, (2) non-linear fusion on fragmented knowledge, and (3) automated demand-driven knowledge navigation. The knowledge graph representation is advocated in BigKE. We also compare BigKE with existing models for Big Data, such as the 4P medical model, the IBM 4V model, 5 R's, and the HACE theorem.