DESIGN AUTOMATION ISSUES AND THEIR SOLUTIONS FOR IMPLEMENTATION OF BIOPROTOCOLS USING ADVANCED MICROFLUIDIC BIOCHIPS

Abstract

Digital and continuous-flow microfluidic biochips are two of the most important platforms revolutionizing point-of-care diagnostics. The automation, high throughput, and low-cost operation features of microfluidic biochips make them suitable for medical diagnosis in remote places. In the last two decades, both digital and continuous-flow microfluidic biochips have evolved a lot and proved to be quite efficient in performing various biochemical applications (bioassays). Also, with the advancement in the fabrication technologies of microfluidic biochips, complex integrated biochips can be easily developed now. However, to efficiently automate a bioassay on such advanced microfluidic biochips, we need sophisticated design automation algorithms too. In the last two decades, researchers have studied and defined various design automation problems related to different types of microfluidic biochips. Along with that, several design automation methodologies have also been developed for efficient bioassay implementation on different microfluidic biochips. This dissertation envisions resolving specific design automation issues related to advanced microfluidic biochips. Based on their working principle, at first, we classify majorly cited microfluidic biochips and then discuss the general architecture, basic fluid manipulation techniques, and also various applications of those biochips. After that, we briefly present the descriptions of various relevant design automation problems and the existing work for different microfluidic biochips to get an overall idea of the nature of those design automation problems. In this work, we have primarily focused on the advanced version of micro-electrode-dot-array (MEDA) biochips, continuous-flow microfluidic biochips (CMBs), and programmable microfluidic devices (PMDs). We take into account multiple design automation issues of these biochips, and propose different solutions for them. The following is a brief discussion of the contributions made by this thesis. In the quest to understand the benefits of the permissible mixing ratios of MEDA biochips in the process of sample preparation, we contributed a few novel sample preparation algorithms for MEDA. In order to resolve single target and multi-demand dilution problems, we present both Abstract heuristic and sub-optimal solutions. The suggested approaches were quite effective at reusing used fluids and reducing the need for reagents. Following that, we examine single target and multi-target mixing issues and present novel methods to reduce fluid wastage. To automate a bioassay on MEDA biochip, we need efficient module placement and routing strategies. One of the primary challenges for such design automation is to minimize the over actuation of electrodes and to maintain the reliability of the biochip. Considering this challenge, we present a reliable module placement and a greedy routing strategy to implement any bioassay on MEDA biochip. In latch-based CMBs, storage cells are associated with a fluid retention time constraint. We frequently need to store fluids and use them later in order to execute a bioassay. Therefore, a storage cell is an important fluidic device for accomplishing any bioassay on a microfluidic biochip. Hence we present a retention time constraint scheduling strategy and also developed a novel approach to obtain a schedulable sample preparation on latch-based CMBs One of the main challenges for efficiently implementing any bioassay on PMDs is to develop an intelligent fluid loading strategy. In this dissertation, we addressed this fluid loading issue and presented a two-phase design strategy. The proposed approach determines efficient fluid loading paths by minimizing the length of the loading paths.