Biomedical Engineering Reference
In-Depth Information
caBIG services are classifi ed into four primary types: Infrastructure/utility
services are those that are required or utilized by virtually all other services
and include semantic services, identity management, security services, and
audit. Core services provide key information components that are utilized in
higher level business capability and business/process services. Examples are
the “COPPA” services that support protocol abstractions (PA), persons (P),
organizations (O), and correlations (C) and services such as diseases (D) and
agents (A). Business capability services provide “business atoms,” data
obtained from core and other services that are utilized by business processes;
examples of these business atoms are models for specimens (S), treatment
plans (Tp), and schedules (Sc). Finally business/process services provide arbi-
trarily complex capabilities that utilize the other three service types to carry
out business functions such as registration (R), outcomes (Po), eligibility (E),
and adverse events (Ae). This list of services is currently under develop-
ment and always evolving; the current set of candidate services and their
specifi cations as well as API specifi cations for those services that already have
reference implementations (eventually created for all caBIG services) are
available from the NCI Services Wiki at https://wiki.nci.nih.gov/display/
EAWiki/Candidate
Services .
As alluded to above, there are a number of usage patterns that can be uti-
lized with caBIG services. At one end of the spectrum an entity may choose to
implement a service specifi cation using their own technology bindings, interop-
erating with the caBIG implementation at either the platform-independent or
the platform-dependent level (depending on the technology chosen). Alterna-
tively, the entity could utilize the reference implementations that are provided
for all caBIG services, integrating that software into their own system and
allowing for interoperability at the platform-specifi c level. Most reference
implementations of caBIG services provide Java, Web services and grid ser-
vices APIs; some also provide support for REST and/or PERL. Finally, the NCI
hosts instances of these services that can be used by groups that do not wish to
deploy their own infrastructure.
+
NCI
+
Enterprise
+
17.5
ca BIG COLLABORATION STRATEGY: SECURITY
Ensuring the security of information (particularly health information) is an
essential part of facilitating collaborations. As a result, security and privacy
have been major concerns of the caBIG program since its inception. As with
all other activities within caBIG, a two-pronged strategy that involves both
technology and sociology was selected to address these concerns. The social
aspects of security and privacy are handled by the DSIC workspace while the
ARCH workspace actually creates the security technology.
The technical implementation of caGrid security is the grid authentication
and authorization with reliably distributed services (GAARDS) framework.
It is shown schematically in Figure 17.2. GAARDS is composed of fi ve primary
Search WWH ::
Custom Search