Biomedical Engineering Reference
In-Depth Information
results in electrode breakdown. When that happens, the chip is permanently
damaged. Therefore, such problems must be avoided during synthesis.
7.2.1.2 Synthesis Guided by Physical Constraints
In this subsection, we describe how the mismatch problems described
earlier can be avoided by incorporating physical constraints in synthesis.
The synthesis method presented in Chapter 2 serves as a starting point for
this approach. To avoid scheduling errors caused by a change in the opera-
tion frequency, we propose to add chip frequency to the design specifica-
tion and expand the module library. The expanded module library will
consist of several sublibraries, with each sublibrary storing the operation
time for all the modules at a certain operation frequency. When an opera-
tion frequency of the target chip is specified, the synthesis tool will choose
the corresponding sublibrary to generate resource binding, scheduling,
and the placement plan.
To avoid the problem of electrode charging, we can add a “successive
activation” time limit as a design rule for synthesis.
Successive activation
is
defined as the number of clock cycles for which the electrode is active. In the
PRSA-based synthesis flow in Chapter 2, for each candidate synthesis result,
we can calculate the maximum successive activation time for each electrode
and add it to the fitness function. Candidate designs with high successive
activation time must be discarded during evolution.
7.2.2 Control-Path Design and Synthesis
The synthesis method of Chapter 2 also suffers from the drawback that it
assumes no control or feedback mechanism during bioassay execution.
Fluidic operations are carried out following the predetermined schedule
without any feedback. Therefore, the only way to ascertain the correctness of
such a synthesized biochip is to examine parameters such as the volume of
the product droplet, sample concentration in the product, detector readout,
etc. If an error is detected at the end of the assay, the entire bioassay must be
repeated. For example, in the protein dilution assay described in Chapter 2,
a sample droplet is diluted by buffer droplets using multiple hierarchies of
binary mixing/splitting phases. If an error occurs in the dispensing reser-
voir and leads to a sample droplet of abnormal volume, the concentrations of
all the product droplets are affected. As a result, the entire assay (133 opera-
tions, including droplet dispensing, mixing/splitting, and detection) must
be reexecuted. Such repetitive executions can potentially lead to wastage of
samples and an undue increase in assay time. Therefore, a monitoring and an
appropriate feedback control mechanism must be implemented. During bio-
assay execution, a monitoring program can determine the status of the assay
and the quality of intermediate products at several checkpoints. If a mal-
function is detected, or the quality of an intermediate product fails to meet
