Biomedical Engineering Reference
In-Depth Information
the amount of data is considerably higher, and therefore designated software and special
computational methods such as complex neuronal networks or support vector machines
(SVMs) (4) have to be used to collect and interpret the results. In addition, the perform-
ance criteria requires validation that are specific to multichannel systems and have to be
clearly defined for the entire set of sensors composing the arrays. These include sensitivity,
specificity, reproducibility, noise, stability, and response time and have to be considered as
a combination of the contributions of individual sensors to the final biosensing device.
Moreover, many multiarray biosensors are fabricated using microfabrication techniques
that may affect the activity of biological systems.
Another important point that has to be considered when working with multiarray sen-
sors is “cross-reactivity,” important when multicomponent mixtures or complex matrices
have to be analyzed. This problem can be solved by carrying out differential measure-
ments using sensors specifically designed for each analyte and subtracting the nonspecific
responses due to the interfering compounds by using appropriate control samples (8). One
of the advantages of using multiarray sensors is the large number of electrodes available
that makes possible designing all the necessary configurations to respond to a diversity of
crossreactive species, while the measurements are performed simultaneously for all the
electrodes under the same experimental variables. This is almost impossible to achieve
with a simple sensor configuration.
Multiarray biosensors could be designed as a combination of multiple sensors that are
able to (i) provide real-time information on the toxicity of multiple samples simultane-
ously and (ii) detect multiple components in a given sample using arrays of sensors specif-
ically designed to sense a particular analyte. The performance of such devices is largely
dependent on the configuration of the sensor array and the detection method used (8-10).
When used in arrays, the reproducibility, selectivity, and sensitivity of individual sensor
are of fundamental importance for obtaining optimum overall system performance. The
configuration of the multiarray sensor is selected in accordance with the transduction sys-
tem used to convert the chemical and biological information on multiple channels and is
strongly dependent upon the number of channels, detection method, and the analytes of
interest. Finally, due to the large amount of variables, computational methods and chemo-
metric analysis have to be used for data processing and quantification (8,10,11). The use of
computational chemistry in combination with multiarray sensors introduces additional
advantages to the final sensing system (e.g., ability to classify and superior data process-
ing). By using appropriate control samples and optimized experimental conditions, a
multiarray biosensor could provide accurate identification, differentiation, and classifica-
tion of analytes. In the context of toxicity monitoring, these systems could provide analyt-
ical information (type and amount) related to the presence of a specific toxicant in the
sample or to a total toxicity level induced by a class of toxicants. Depending on the system
used and its performance, this information can be quantitative, semiquantitative, or
simply qualitative by providing a “yes or no” answer (3). A schematic representation of a
multiarray (bio)sensor including detection, data processing, and quantification is shown
in Figure 19.1.
The most preferred arrangement for a multiarray sensor is a 96-well electrode system,
but sensors composed of 4, 16, or 32 electrodes have also been reported. These sensing sys-
tems can be designed as a combination of multiple sensors deposited on a single-sensor
substrate to form an array or as an array of individual sensors arranged in a multiarray
format. Figure 19.2 shows examples of three sensor arrays used in our laboratory. The first
one (A) is a laboratory-made sensor array comprising six sensors deposited onto the same
chip. In such a configuration, each sensor can be specifically tailored toward a particular
analyte so that the final device will allow the detection of multiple analytes in a particular
sample. The second (B) and the third (C) are produced by Daikin Corporation, Japan and
