Approaches to medical decision-making based on big clinical data

This paper by Malykh and Rudetskiy “discusses different approaches to building a clinical decision support system based on big data,” with a focus on non-biased processing methods and their comparative assessments. After an in-depth analysis of methods and objectives, the authors present their findings from the clinical decision support data and their significance. They conclude that case-based and precedent-based approaches each have their advantages–including more accurate recommendations and faster system speeds–but are not without disadvantages. The authors suggest future research is needed to address “problems with optimization of provided metrics, compression of state descriptions, and construction of training procedures.”

A new numerical method for processing longitudinal data: Clinical applications

When it comes to longitudinal data, what analysis methods are we using today? How can they be applied to clinical data? In this 2018 paper, Stura et al. look at, for example, repeated data from measuring patient reactions and behaviors to a therapy. yet when analyzing this type of data problems arise; “more robust statistical methods” are required. The authors combine several methods to develop a “numerical tool based on optimization methods coupled with interpolation techniques.” They conclude that it provides several benefits, including output displayed as “a (continuous) growth curve, allowing the analysis of each growth function independently of the others. “

Big data management for healthcare systems: Architecture, requirements, and implementation

In this 2018 paper, El aboudi and Benhilma discuss the data management architectures of healthcare from the perspective of Northern Africa. As part of their discussion, the authors propose “an extensible big data architecture based on both stream computing and batch computing in order to enhance further the reliability of healthcare systems by generating real-time alerts and making accurate predictions on patient health condition.” With such an architecture, they conclude that, when implemented well, the healthcare system may be “capable of handling the high amount of data generated by different medical sources in real time.”

CÆLIS: Software for assimilation, management, and processing data of an atmospheric measurement network

In this 2018 paper published in Geoscientific Instrumentation, Methods and Data Systems, Fuertes et al. describe and demonstrate the uses of CÆLIS, software designed to simplify the processing, management, and use of atmospheric particulate data. After describing the architecture and database model, the authors describe its functionality and real-world applications. The authors conclude that the automation the software brought to aerosol measurement and analysis has been significant, in that software “has reduced the number of human errors and allowed one to perform more in-depth and exhaustive analysis” while also allowing users to “perform queries and extract data in a fast and very flexible way.”

Support Your Data: A research data management guide for researchers

In this 2018 paper published in Research Ideas and Outcomes, Borghi et al. of the University of California Curation Center discuss their suite of research data management (RDM) tools, Support Your Data. The tools “include a rubric designed to enable researchers to self-assess their current data management practices and a series of short guides which provide actionable information about how to advance practices as necessary or desired.” Based on three key RDM trends, the researchers felt a need to provide a complementary set of tools for researchers to better address those trends. The conclude by offering several use cases for the tools and planning “next steps” for improving the tools.

Big data in the era of health information exchanges: Challenges and opportunities for public health

Baseman et al. “conducted an assessment of big data that is available to a [public health agency]—laboratory test results and clinician-generated notifiable condition report data—through its participation in a [health information exchange]” and published their results in Informatics. They identified five major challenges to “secondary use of HIE data for meeting public health communicable disease surveillance needs” and then find ways to turn those challenges into opportunities for the public health system, ultimately optimizing it through various forms of big data analysis and management.

How could the ethical management of health data in the medical field inform police use of DNA?

This brief article published in Frontiers in Public Health takes a look at the collective management of genetic analysis techniques and “the ethico-legal frameworks” associated with forensic science and biomedicine. Krikorian and Vailly introduce the ethics and data collection methods of genetic material by police, noting how with new techniques the ethics have changed. Then they discuss the legal and political ramifications that go along with the ethics. They conclude that questions persist “about the conditions for the existence or for the absence of political controversies that call for further sociological investigations about the framing of the issue and the social and political logic at play.” Additionally, they note “the need for promoting dialogue among the various professionals using this technology in police work” as well as “with healthcare professionals.”

Promoting data sharing among Indonesian scientists: A proposal of a generic university-level research data management plan (RDMP)

In this 2018 paper by Irawan and Rachmi, a proposed research data management plan (RDMP) for the university ecosystem is proposed. After introducing the concept of an RDMP and the layout, the authors describe seven major components to their plan, in the form of an assessment form: data collection; documentation and metadata; storage and backup; preservation; sharing and re-use; responsibilities and resources; and ethics and legal compliance. They conclude that the assessment form can help researchers “to describe the setting of their research and data management requirements from a potential funder … [and] also develop a more detailed RDMP to cater to a specific project’s environment.”

systemPipeR: NGS workflow and report generation environment

In this 2016 paper published in BMC Bioinformatics, Backman and Girke discuss the R/Bioconductor package systemPipeR. Recognizing that the analysis of next-generation sequencing (NGS) data remains a significant challange, the authors turned to the R programming language and the Bioconductor environment to make workflows that were “time-efficient and reproducible.” After giving some background and then discussing the development and implementation, they conclude that systemPipeR helps researchers “reduce the complexity and time required to translate NGS data into interpretable research results, while a built-in reporting feature improves reproducibility.”

A data quality strategy to enable FAIR, programmatic access across large, diverse data collections for high performance data analysis

A data quality strategy (DQS) is useful for researchers, organizations, and others, primarily because it allows them “to establish a level of assurance, and hence confidence, for [their] user community and key stakeholders as an integral part of service provision.” Evans et al. of the Australian National University, recognizing this importance, discuss the implementation of their DQS at the Australian National Computational Infrastructure (NCI), detailing their strategy and providing examples in this 2017 paper. They conclude that “[a]pplying the DQS means that scientists spend less time reformatting and wrangling the data to make it suitable for use by their applications and workflows—especially if their applications can read standardized interfaces.”

How big data, comparative effectiveness research, and rapid-learning health care systems can transform patient care in radiation oncology

In this brief opinion article published in Frontiers in Oncology, Sanders and Showalter turn their thoughts to the rapid-learning health care system (RLHCS), a concept that involves analyzing patient data to make insights in how to improve patient safety, treatment quality, and cost-effectiveness within the health care framework. In combination with comparative effectiveness research (CER) and well-managed big data streams, the RLHCS has the power “to accelerate discovery and the future of individualized radiation treatment planning,” they argue. They conclude that big data can “connect a broad range of characteristics to accelerate evidence generation and inform personalized decision-making,” and its application through CER and the RLHCS can “accelerate progress in cancer care.”

Wireless positioning in IoT: A look at current and future trends

In this 2018 paper published in Sensors, e Silva et al. “highlight the importance of positioning features for [internet of things] applications and [provide] means of comparing and evaluating different connectivity protocols in terms of their positioning capabilities.” Noting a lack of research on the topic of IoT connectivity solutions and how they are localized and tracked, the researchers present what related work there is on the topic, discuss positioning domains and systems, compare IoT technologies and enablers., and provide several case studies. They conclude “that power-domain positioning currently offers the best trade-off between implementation cost and positioning accuracy for low-power systems.”

Password compliance for PACS work stations: Implications for emergency-driven medical environments

Of course, keeping personal health information protected is important, but what challenges exist to this point in the highly busy and time-sensitive setting of the emergency medical setting? At the end of 2017 Mahlaola and van Dyk published an article on this topic in the South African Journal of Bioethics and Law that “argues that the minimum standards of effective password use prescribed by the information security sector are not suitable to the emergency-driven medical environment, and that their application as required by law raises new and unforeseen ethical dilemmas. ” Using the picture archiving and communication system (PACS) as the primary focus, the authors collected survey responses from Johannesburg-based hospital and radiology departments. After analysis and discussion, they conclude that indeed some ethical quandary exists in the fight to protect patient data using passwords while also trying to save lives, particularly in settings where “seconds count.”

Data science as an innovation challenge: From big data to value proposition

How best can we retrieve value from the rich streams of data in our profession, and introduce a solid, systematic process for analyzing that data? Here Kayser et al. describe such a process from the perspective of data science experts at Ernst & Young, offering a model that “aims to structure and systematize exploratory analytics approaches.” After discussing the building blocks for value creation, they suggest a thorough process of developing analytics approaches to data analytics. They conclude that “[t]he process as described in this work [effectively] guides personnel through analytics projects and illustrates the differences to known IT management approaches.”

The development of data science: Implications for education, employment, research, and the data revolution for sustainable development

In this 2018 paper by Murtagh and Devlin, a historical and professional perspective on data science and how collaborative work across multiple disciplines is increasingly common to data science. This “convergence and bridging of disciplines” strengthens methodology transfer and collaborative effort, and the integration of data and analytics guides approaches to data management. But education, research, and application challenges still await data scientists. The takeaway for the authors is that “the importance is noted of how data science builds collaboratively on other domains, potentially with innovative methodologies and practice,”

GeoFIS: An open-source decision support tool for precision agriculture data

Scientists everywhere nod to the power of big data but are still left to develop tools to manage it. This is just as true in the field of agriculture, where practitioners of precision agriculture are still developing tools to do their work better. Leroux et al. have been developing their own solution, GeoFIS, to better handle geolocalized data visualization and analysis. In this 2018 paper, they use three case studies to show off how GeoFIS visualizes data, processes it, and incorporates associated data (metadata and industry knowledge) for improved agricultural outcomes. They conclude that the software fills a significant gap while also promoting the adoption of precision agriculture practices.

Technology transfer and true transformation: Implications for open data

In this 2017 paper published in the Data Science Journal, University of Oxford’s Louise Bezuidenhout makes a case for how local challenges with laboratory equipment, research speeds, and design principles hinder adoption of open data policies in resource-strapped countries. Noting that “openness of data online is a global priority” in research, Bezuidenhout uses interviews in various African countries and corresponding research to draw conclusions that many in high-income countries may not. The main conclusion: “Without careful and sensitive attention to [the issues stated in the paper], it is likely that [low- and middle-income country] scholars will continue to exclude themselves from opportunities to share data, thus missing out on improved visibility online.”

Eleven quick tips for architecting biomedical informatics workflows with cloud computing

In this educational journal article published in PLoS Computational Biology, Cole and Moore of the University of Pennsylvania’s Institute for Biomedical Informatics offer 11 tips for health informatics researchers and practitioners to embrace in improving reproducibility, knowledge sharing, and costs: adopt cloud computing. The authors compare more traditional “in-house enterprise compute systems such as high-performance computing (HPC) clusters” located in academic institutions with more agile cloud computing installations, showing various ways researchers can benefit from building biomedical informatics workflows on the cloud. After sharing their tips, they conclude that “[c]loud computing offers the potential to completely transform biomedical computing by fundamentally shifting computing from local hardware and software to on-demand use of virtualized infrastructure in an environment which is accessible to all other researchers.”

Welcome to Jupyter: Improving collaboration and reproduction in psychological research by using a notebook system

Jupyter Notebook, an open-source interactive web application for the data science and scientific computing community (and with some of the features of an electronic laboratory notebook), has been publicly available since 2015, helping scientists make computational records of their research. In this 2018 tutorial article by Friedrich-Schiller-Universität Jena’s Phillipp Sprengholz, the installation procedures and features are presented, particularly in the context of aiding psychological researchers with their efforts in making research more reproducible and shareable.

Developing a file system structure to solve healthcare big data storage and archiving problems using a distributed file system

There’s been plenty of talk about big data management over the past few years, particularly in the domain of software-based management of said data. But what of the IT infrastructure, particularly in the world of heathcare, where file size and number continue to grow? Ergüzen and Ünver describe in this 2018 paper published in Applied Sciences how they researched and developed a modern file system structure that handles the intricacies of big data in healthcare for Kırıkkale University. After discussing big data problems and common architectures, the duo lay out the various puzzle pieces that make up their file system, reporting system performance “97% better than the NoSQL system, 80% better than the RDBMS, and 74% better than the operating system” via improvements in read-write performance, robustness, load balancing, integration, security, and scalability.