Home
This Title All WIREs
WIREs RSS Feed
How to cite this WIREs title:
WIREs Data Mining Knowl Discov
Impact Factor: 2.111

An overview of online based platforms for sharing and analyzing electrophysiology data from big data perspective

Full article on Wiley Online Library:   HTML PDF

Can't access this content? Tell your librarian.

With the development of applications and high‐throughput sensor technologies in medical fields, scientists and scientific professionals are facing a big challenge—how to manage and analyze the big electrophysiological datasets created by these sensor technologies. The challenge exhibits several aspects: one is the size of the data (which is usually more than terabytes); the second is the format used to store the data (the data created are generally stored using different formats); the third is that most of these unstructured, semi‐structured, or structured datasets are still distributed over many researchers' own local computers in their laboratories, which are not open access, to become isolated data islands. Thus, how to overcome the challenge and share/mine the scientific data has become an important research topic. The aim of this paper is to systematically review recent published research on the developed web‐based electrophysiological data platforms from the perspective of cloud computing and programming frameworks. Based on this review, we suggest that a conceptual scientific workflow‐based programming framework associated with an elastic cloud computing environment running big data tools (such as Hadoop and Spark) is a good choice for facilitating effective data mining and collaboration among scientists. WIREs Data Mining Knowl Discov 2017, 7:e1206. doi: 10.1002/widm.1206

Challenges of a web‐based electrophysiological data sharing and analysis platform.
[ Normal View | Magnified View ]
Typical functional architecture of scientific workflow management systems.
[ Normal View | Magnified View ]
Electrophysiological data processing in Hadoop and Spark. (a) MapReduce data flow in Hadoop. (b) In‐memory data processing in Spark.
[ Normal View | Magnified View ]
HDF5 data model. This model includes abstract data and storage models (the data format), and libraries to implement the abstract model and map the storage model to different storage mechanisms.
[ Normal View | Magnified View ]
Timeline of data formats for fourteen platforms. The blue line indicates the appearance of new text formats, and the red line indicates the appearance of new binary formats. (The characters in parentheses are the annotation data format and metadata format; the labels beneath the underline are the names of the platforms.)
[ Normal View | Magnified View ]
Pipeline of online big data sharing and analysis workflow.
[ Normal View | Magnified View ]
Functional architecture of an electrophysiology data sharing platform.
[ Normal View | Magnified View ]
Visualization interfaces of International Epilepsy Electrophysiology (IEEG) and Code Analysis Repository and Modelling for E‐Neuroscience (CARMEN). (a) Visualizing multichannel signals in the IEEG‐Portal. The pane includes start time and window width settings, channel selection, filter settings, remontaging and annotating data. (b) Analysis workflow of CARMEN (http://www.carmen.org.uk/?page_id=316). The demo workflow includes an input file, a service, and an output folder. The connections can be created by simple drag and drop from either of the connectors to the other.
[ Normal View | Magnified View ]

Browse by Topic

Application Areas > Health Care
Fundamental Concepts of Data and Knowledge > Information Repositories
Technologies > Computer Architectures for Data Mining

Access to this WIREs title is by subscription only.

Recommend to Your
Librarian Now!

The latest WIREs articles in your inbox

Sign Up for Article Alerts