Optimization of Multi-Target Sample Preparation On-Demand With Digital Microfluidic Biochips

Published

Journal Article

© 2018 IEEE. Sample preparation is a fundamental preprocessing step needed in almost all biochemical assays and is conveniently automated on a microfluidic lab-on-chip. In digital microfluidics, it is accomplished by a sequence of droplet-mix-split steps on a biochip. Many real-life applications require a sample with multiple concentration factors (CFs). Existing algorithms, while producing multi-CF targets, attempt to share the mix-split steps in order to reduce reactant-cost and sample-preparation time. However, all prior approaches have two limitations: 1) sharing of intermediate droplets can be best effected only when all required target CFs are known a priori and 2) the processing time may vary depending on the allowable error-tolerance in target-CFs. In this paper, we present a cost-effective solution to multi-CF-dilution on-demand, by using only one (or two) mix-split step(s). In order to service dynamically arriving requests of multiple CFs quickly, we prepare dilutions of the sample with a few CFs in advance (called source-CFs), and fill on-chip reservoirs with these fluids. For minimizing the number of such preprocessed CFs, we present an integer linear programming-based method, an approximation algorithm, and a heuristic algorithm. The proposed methods also allow the users to tradeoff the number of on-chip reservoirs against service time for various applications. Simulation results for several target sets demonstrate the superiority of the proposed techniques over prior art in terms of the number of mix-split steps, waste droplets, and reactant usage when the on-chip reservoirs are preloaded with source-CFs using a customized droplet-streaming engine.

Full Text

Duke Authors

Cited Authors

  • Poddar, S; Bhattacharjee, S; Nandy, SC; Chakrabarty, K; Bhattacharya, BB

Published Date

  • February 1, 2019

Published In

Volume / Issue

  • 38 / 2

Start / End Page

  • 253 - 266

International Standard Serial Number (ISSN)

  • 0278-0070

Digital Object Identifier (DOI)

  • 10.1109/TCAD.2018.2808234

Citation Source

  • Scopus