Information Technology Reference
In-Depth Information
Variability
: the syntactic and/or semantic nature of the data is highly variable,
requiring specialized knowledge management and engineering approaches in
order to support the analysis of such resource
It is widely held that a data set or resource demonstrating any one or more of the
aforementioned characteristics is “Big Data.” As can readily be concluded, any num-
ber of data types commonly encountered in the modern biomedical environment can
be classifi ed as “Big Data”, such as patient-derived phenotypes extracted from
EHRs, sensor data used to understand patient or population characteristics outside of
the clinical care environment, and the data resulting from modern bio-molecular
instrumentation such as that associated with exome or whole genome sequencing.
The primary challenges that have prompted the defi nition of what is (and is not)
“Big Data”, and that have catalyzed such widespread interest in “Big Data” analyt-
ics, include [
17
]:
The absence of well accepted and/or validated tools and methods capable of
reliably and effi ciently supporting or enabling the collection, storage, transaction,
and analysis of “Big Data” in a timely and cost-effective manner (e.g., not requir-
ing specialized, costly, and time-intensive computational tools and approaches)
;
The minimal understanding of how common quantitative science measurements,
such as conventional statistical signifi cance testing, scale to indicate and
quantify patterns or phenomena of interest in “Big Data” constructs, especially
for those that exhibit two or more of the “3Vs” and thus could include consider-
able sparsity or “noise”
;
and
The innate challenges of delivering “Big Data” to human end-users in a manner
that is comprehensible and capable of leveraging innate cognitive capabilities
relative to higher-order pattern recognition and semantic reasoning.
Because of these, and many other compelling computational, quantitative sci-
ence, and Biomedical Informatics challenges associated with “Big Data”, this area
has emerged as a rapidly growing and dynamic area of research and investigation,
likely to greatly infl uence and contribute to the overall TI vision introduced here.
1.4
A Path Forward for Translational Informatics
and Knowledge-Based Healthcare
As can be ascertained by the preceding survey of the current state of biomedical
knowledge and practice, and the major factors and trends that we have emphasized,
it can be argued that the realization of a TI vision is beset by signifi cant challenges
and opportunities. As such, we will contend in subsequent chapters of this topic that
there are three major areas that must be addressed in order to advance the TI
knowledge-base from both a basic and applied perspective, namely: (1) the pursuit
of strategic, TI-relevant strategic and research foci; (2) an increased emphasis on the
Search WWH ::
Custom Search