Biomedical Engineering Reference
In-Depth Information
on the operation completion time. While generating physical layouts, the
synthesis tool in [15] provides only the layouts of the modules, and it leaves
droplet-routing pathways unspecified. The assumption of negligible droplet
transportation times is valid for small microfluidic arrays. However, for large
arrays and for biochemical protocols that require several concurrent fluidic
operations on chip, the droplet transportation time is significant and rout-
ing complexity is nontrivial.
Recent work on automated biochip design has also included postsynthe-
sis droplet routing [26,27]. These methods can reduce droplet transportation
time by finding optimal routing plans for a synthesized biochip. However,
the effectiveness of such methods is limited by the synthesis results; that is,
the placement of microfluidic modules often determines the droplet pathways
that lead to minimum droplet transportation time. For example, if we need
to route a droplet between two modules that are 10 electrodes away from
each other, then it is not possible to reduce the droplet transportation time to
less than that needed to move a droplet by a distance equal to 10 electrodes.
Since droplet pathways are dynamically reconfigurable, the number of feasi-
ble droplet pathways can be very high, leading to considerable computation
time for a droplet-routing tool.
The testing of microfluidic biochips has recently been investigated [28-30].
These test methods add fluid-handling aspects to MEMS testing techniques
[31]. Test methods have been proposed for both continuous-flow and digital
microfluidic biochips. An excellent review is available in [32]. A fault model
and a fault simulation method for continuous-flow microfluidic biochips
have been proposed in [33]. For digital microfluidic chips, techniques for
defect classification, test planning, and test resource optimization have been
presented in [28]. Defect-classification methods are discussed in [28], and the
corresponding test procedures are described in [29]. Defects have been clas-
sified as being either catastrophic or parametric, and techniques have been
developed to detect these defects by electrostatically controlling and track-
ing droplet motion.
The work in [28,29] facilitates concurrent testing, which allows fault detec-
tion and biomedical assays to run simultaneously on a microfluidic sys-
tem. A drawback of [28], however, is that it does not present any automated
techniques for optimizing the test application procedure. Reference [34]
first proposed a test-planning and test resource optimization method. The
test-planning problem is mapped to the Hamilton cycle problem from graph
theory. An alternative method based on Euler paths is proposed in [36]. This
method maps a digital microfluidic biochip to an undirected graph, and a
test droplet is routed along the Euler path derived from the graph to pass
through all the cells in the array. Fault diagnosis is carried out using multiple
test application steps and adaptive Euler paths.
Another important issue in biochip design is electrode addressing, that
is, the manner in which electrodes are connected to and controlled by input
pins. Early design automation techniques relied on the availability of a
