R2D2: A scalable deep learning toolkit for medical imaging segmentation

Deep learning has gained a significant popularity in recent years thanks to its tremendous success across a wide range of relevant fields of applications, including medical image analysis domain in particular. Although convolutional neural networks (CNNs) based medical applications have been providing powerful solutions and revolutionizing medicine, efficiently training of CNNs models is a tedious and challenging task. It is a computationally intensive process taking long time and rare system resources, which represents a significant hindrance to scientific research progress. In order to address this challenge, we propose in this article, R2D2, a scalable intuitive deep learning toolkit for medical imaging semantic segmentation. To the best of our knowledge, the present work is the first that aims to tackle this issue by offering a novel distributed versions of two well-known and widely used CNN segmentation architectures [ie, fully convolutional network (FCN) and U-Net]. We introduce the design and the core building blocks of R2D2. We further present and analyze its experimental evaluation results on two different concrete medical imaging segmentation use cases. R2D2 achieves up to 17.5× and 10.4× speedup than single-node based training of U-Net and FCN, respectively, with a negligible, though still unexpected segmentation accuracy loss. R2D2 offers not only an empirical evidence and investigates in-depth the latest published works but also it facilitates and significantly reduces the effort required by researchers to quickly prototype and easily discover cutting-edge CNN configurations and architectures.     

A hybrid multi-criteria decision-making method for cloud adoption: Evidence from the healthcare sector

The purpose of this paper is to investigate the main determinants (criteria as well as sub-criteria) influencing the adoption decision of cloud computing (CC) in the healthcare sector. In the present study, qualitative interviews and the Delphi method are used to find the relevance of identified determinants (criteria and sub-criteria) from literature. Further, an integrated hybrid approach of interpretive structural modeling, analytic hierarchy process, and Technique for Order Preference by Similarity to Ideal Solution (ISM-AHP-TOPSIS) have been employed to identify interrelationships among criteria, rank critical criteria as well as subcriteria, and find the most suitable cloud service provider (CSP) respectively. This study analyses seven criteria and twenty-one sub-criteria that offers a roadmap to decision-makers before CC-adoption (CCA). The analysis of ISM and AHP revealed technology as the most critical and human-environment as the least critical criteria. The vital sub-criteria associated with each criterion are Management Strategic Planning (under Organisation), Government Policy (External Environment), Security (Technology), Financing of IT in Hospital (Economic), Perceived Visibility (Social), and Attitude of Colleagues about Cloud Computing (Human Environment). The results from TOPSIS revealed CSP5 being the most suitable service provider followed by CSP2. Finally, this research makes methodological contribution in terms of an integrated hybrid method, to select a suitable CSP for the healthcare sector and the theoretical contribution in terms of criteria & sub-criteria. This article answers an important question, i.e., whether the benefits of CCA outweighs its barriers, thereby exploring the potential (and future) of CC for the advancement of healthcare provision.     

云计算在智慧医院信息服务中的应用

随着医疗服务体系的发展和进步,越来越多的医院开始将云计算技术引入信息服务系统,使医院信息化建设得到了迅速发展。文章分析了云计算技术的特点及其在智能医院信息服务中的应用和影响,以期促进该技术在未来的发展和推广。     

Should medium-sized contractors still implement home-made information technologies on construction sites?

Despite the increasing dissemination of the Building Information Modeling (BIM) approach, many firms are still far from BIM and are only starting to adopt information technologies. We conducted an ethnographic action research with an industrial partner that implements information technologies on its construction sites, with a strong preference for home-made applications. Results show that the development of such applications faces difficulties and that multiple iterations and versions are necessary before achieving the expected goals because the required competencies are lacking among internal development teams. Moreover, initial needs evolve considerably throughout the application’s development and evaluation process. However, notwithstanding the difficulties encountered during the different application development phases, the industrial partner believes the decision to be worthwhile because of the desire to have internal control over the project data and the need to consider certain specificities related to business processes.     

Bioinspired interactive neuromorphic devices

The performance of conventional computer based on von Neumann architecture is limited due to the physical separation of memory and processor. By synergistically integrating various sensors with synaptic devices, recently emerging interactive neuromorphic devices can directly sense/store/process various stimuli information from external environments and implement functions of perception, learning, memory, and computation. In this review, we present the basic model of bioinspired interactive neuromorphic devices and discuss the performance metrics. Next, we summarize the recent progress and development of bioinspired interactive neuromorphic devices, which are classified into neuromorphic tactile systems, visual systems, auditory systems, and multisensory system. They are discussed in detail from the aspects of materials, device architectures, operating mechanisms, synaptic plasticity, and potential applications. Additionally, the bioinspired interactive neuromorphic devices that can fuse multiple/mixed sensing signals are proposed to address more realistic and sophisticated problems. Finally, we discuss the pros and cons regarding to the computing neurons and integrating sensory neurons and deliver the perspectives on interactive neuromorphic devices at the material, device, network, and system levels. It is believed the neuromorphic devices can provide promising solutions to next generation of interactive sensation/memory/computation toward the development of multimodal, low-power, and large-scale intelligent systems endowed with neuromorphic features.     

Parallel Multipopulation Differential Evolutionary Particle Swarm Optimization for Voltage and Reactive Power Control

This paper presents parallel multipopulation differential evolutionary particle swarm optimization (DEEPSO) for voltage and reactive power control (VQC). The problem can be formulated as a mixed integer nonlinear optimization problem and various evolutionary computation techniques have been applied to the problem including PSO, differential evolution (DE), and DEEPSO. Since VQC is one of the online controls, speed‐up of computation is required. Moreover, there is still room for improvement in solution quality. This paper applies parallel multipopulation DEEPSO in order to speed up the calculation and improve solution quality. The proposed method is applied to IEEE 30, 57, and 118 bus systems. The results indicate that the proposed method can realize fast computation and minimize more active power losses than the conventional evolutionary computation techniques.     

基于一种云计算数据保护的多级加密算法的应用研究

云计算是分布式计算技术的一种,其最基本的概念是透过网络将庞大的计算处理程序自动分拆成无数个较小的子程序,再交由多部服务器所组成的庞大系统经搜寻、计算分析之后将处理结果回传给用户。云计算中的数据泄漏已经成为云计算的巨大威胁,据不完全统计数据,仅在2019年1年,大约有超40亿条信息被泄露。这些信息涵盖科技、医疗、政府、金融、教育、制造等十几个领域数百个行业,数据泄露已经成为云计算中数据安全的巨大威胁。主要提出了一种多级加密算法,能更好地保护云计算中的数据,并能以极小的成本在云计算中得到应用。     

Emotion recognition at the edge with AI specific low power architectures

Nowadays Deep Learning is applied in almost every research field and helps getting amazing results in a great number of challenging tasks. The main problem is that this kind of learning and consequently Neural Networks that can be defined deep, are resource intensive. They need specialized hardware to perform computation in a reasonable time. Many tasks are mandatory to be as much real-time as possible . It is needed to optimize many components such as code, algorithms, numeric accuracy and hardware, to make them “efficient and usable”. All these optimizations can help us to produce incredibly accurate and fast learning models. The paper reports a study in this direction for the challenging face detection and emotion recognition tasks.     

An Advanced Analysis of Cloud Computing Concepts Based on the Computer Science Ontology

Our primary research hypothesis stands on a simple idea: The evolution of top-rated publications on a particular theme depends heavily on the progress and maturity of related topics. And this even when there are no clear relations or some concepts appear to cease to exist and leave place for newer ones starting many years ago. We implemented our model based on Computer Science Ontology (CSO) and analyzed 44 years of publications. Then we derived the most important concepts related to Cloud Computing (CC) from the scientific collection offered by Clarivate Analytics. Our methodology includes data extraction using advanced web crawling techniques, data preparation, statistical data analysis, and graphical representations. We obtained related concepts after aggregating the scores using the Jaccard coefficient and CSO Ontology. Our article reveals the contribution of Cloud Computing topics in research papers in leading scientific journals and the relationships between the field of Cloud Computing and the interdependent subdivisions identified in the broader framework of Computer Science.     

Real-scale DEM simulations on the fault evolution process observed in sandbox experiments

The study of fault structures and stress states in accretionary prisms is important to elucidate the building and releasing of seismic energy as they control the generation of great earthquakes and tsunami. In this paper, we present the evolution process of three-dimensional fault structures performed in sandbox simulations using a discrete element method (DEM). To realize a real-scale sandbox simulation, we developed state-of-the-art techniques in high performance parallel computing for the DEM and performed the world’s largest DEM simulation using up to 1.9 billion particles with a similar grain size as real sand to identify the three-dimensional fault structure. The DEM simulations reproduced the undulation of fault structures, similar to those commonly found in nature. In addition, the characteristic grain motion was observed near the frontal fault before the commencement of the uplift event of the sand bed, which could be a precursor of tectonic events behind accretionary prism formation.