Biomedical Engineering Reference
In-Depth Information
In general it is not a single technology or specific material that makes a biosensor smart
or intelligent, but the synergistic interaction between the constituent components that
comprise the sensory system. Seamless functional integration requires the development
team to understand fundamental scientific and design principles derived from chemistry,
physics, biology, material science, electronics and optics, and informatics. These seemingly
diverse branches of science, engineering, and information science all provide unique per-
spectives on the meaning of “intelligent biosensors,” their design, and potential applica-
tions. To further increase a sensor's specificity, selectivity, and overall functionality many
commercial biosensors must incorporate the computing power of microprocessors and
microcontrollers. These information-intensive biosensor systems exploit numeric algo-
rithms and new analytical techniques being developed in areas of adaptive signal pro-
cessing, machine learning, pattern recognition, and artificial intelligence (AI).
The field of intelligent biosensors has borrowed terms, concepts, and technologies from
a variety of scientific and engineering disciplines. Consequently, the words
smart
and
intel-
ligent
can mean different things to different groups of researchers (4). For example, the IEEE
1451 smart transducer interface standard defines “smartness” in terms of on-board data
storage/processing capability and the existence of an interfaced/integrated analog or dig-
ital sensor (5). Clearly, this is a very narrow interpretation of a smart device. In other areas
of electrical engineering, a smart device is defined as a system with integrated computing
power that allows self-calibration, nonlinear correction, offset elimination, failure detection,
communication, and decision-making ability (6,7). A general definition of a smart or intel-
ligent sensor is a sensing device that includes signal-processing capability integrated into it
to reduce physical deficiencies in the intrinsic hardware. The signal processing can be either
analog or digital. Intelligence can result in the raw signal data from the sensor being sig-
nificantly transformed or even reconfigured as meaningful information.
The terminology used in biosensor technology is further influenced by the disciplines of
chemistry and material science. Biosensor systems comprised of biological macromolecules,
assemblies of these molecules, and living cells exhibit some intelligent properties. Most of
these smart biosensors are created using highly specific materials. The materials are termed
intelligent
because they are selective and respond to only the desired receptor signal.
Intelligent properties include template-based self-assembly, self-multiplication, self-repair,
self-degradation (selective), redundancy, self-diagnosis, learning, and prediction/notifica-
tion. The basic philosophy is that intelligent material properties incorporated into biosensors
would enable the systems to respond in real time to environmental changes and be capable
of integrating multiple functions such as signal feedback and pattern.
A common feature of all smart materials and smart sensors is
adaptation
. This may be in
the form of adapting to environmental stimuli, adaptive signal processing algorithms,
adaptive control, and intelligent decision-making. Although several distinct elements com-
prise a sensor system, it is when these elements are effectively integrated and function col-
lectively that the sensor truly exhibits “smartness.” The integrated structure will perform
in a way that is similar to biological systems, adapting to the circumstances and exploiting
available energy as efficiently as possible. As the system becomes more complex, the capa-
bilities of the sensor are increased with the inclusion of structural condition monitoring and
health maintenance. The sensor must act in some way to optimize or improve its perform-
ance. The fundamental mechanism is the inherent ability to adjust its internal parameters
to changing environmental conditions. The capability of modifying behavior to external
influences is a defining attribute of many living cells or all biological organisms.
At the higher organizational level, adaptive sensor signal processing and sophisticated pat-
tern recognition exist. The ability to acquire diverse sensory inputs, extract relevant features,
and rapidly comprehend the underlying meaning of the stimuli is the key to a truly smart
biosensor. Similar to a living organism, these attributes cannot be realized in a biosensor by
