Design Automation and Test Techniques for Microfluidic Biochips

Microfluidics-based biochips, also referred to as lab-on-a-chip (LoC), are devices that integrate fluid-handling functions such as sample preparation, analysis, separation, and detection. This emerging technology combines electronics with biology to open new application areas including point-of-care diagnosis, on-chip DNA analysis, and automated drug discovery. As digital microfluidic biochips are developed for high throughput and concurrent assays, biochip-specific design automation and testing techniques are needed to handle increased complexity. In this paper, we address three important problems in the automation of design and testing of digital microfluidic biochips. First, we focus on the droplet routing problem, which is a key issue in biochip physical design automation. We develop the first systematic droplet routing method that can be integrated with biochip synthesis [1]. The proposed approach attempts to minimize the number of cells used for droplet routing, while satisfying constraints imposed by throughput considerations and fluidic properties. Next, we describe recent experiments that reveal the inadequacy of testing based on Hamiltonian paths in a graph model of the microfluidic array and introduce a novel testing methodology for localizing faults. We show previous approaches of mapping the droplet flow path problem to that of finding a Hamiltonian path in a graph model of the array is not sufficient to detect electrode-short and fluidic-open faults that affect two adjacent electrodes. To detect these edge-dependent defects, we developed testing methods based on adaptations of the Euler circuit and Euler path using a probabilistic-modified Fleury’s algorithm to find efficient ways to detect and localize faults by traversing all edges of an undirected graph exactly once [2]. Finally, we propose a design automation method for pinconstrained biochips. In contrast to the direct-addressing scheme that has been studied thus far in the literature, we assign a reduced number of independent control pins to a large number of electrodes in the LoC, thereby reducing design complexity and product cost. We develop a virtual partitioning scheme that can prevent multi-droplet interference and maintain throughput for most practical assays on pin-constrained biochips [3]. We demonstrate the proposed method on a set of multiplexed bioassays. 1 This research was supported by the National Science Foundation under grant number IIS-0312352 and a Research Experiences for Undergraduates (REU) grant.