Issue No. 04 - July/August (2007 vol. 24)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/MDT.2007.123
Sachin Sapatnekar , University of Minnesota
Biosystems and applications are hot items today. Lab-on-a-chip (LoC) systems constitute a multibillion-dollar business, with sensors for monitoring blood glucose providing early success stories. Is there a natural fit for chip design techniques in this field? The Jan.-Feb. 2007 issue of IEEE Design & Test gave readers a first glimpse of this growing area. Two recent books, Design Automation Methods and Tools for Microfluidics-Based Biochips and BioMEMS, provide deeper insight, but in a way that will be familiar to many of us in the chip design world. Both books are edited volumes, with contributions from several authors on topics related to their areas of expertise. Together, they provide a good introduction to the field of biochips. For D&T readers, Design Automation Methods and Tools for Microfluidics-Based Biochips is arguably the more accessible of the two, and this was the one I read first. However, after reading that book, I read BioMEMS, which focuses on modeling and design issues, and I found it to be quite lucid. It reinforces and supplements several ideas discussed in the first book, but goes beyond LoC systems to other types of biomedical microelectromechanical systems (bioMEMS), such as neural implants.
Design Automation Methods and Tools for Microfluidics-Based Biochips
This book deals exclusively with microfluidics-based biochips, also known as LoC systems. The authors begin the book with a chapter that describes the motivation for the application domain: These miniaturized devices eliminate repetitive laboratory tasks by manipulating biochemical samples for applications such as genomics, proteomics, clinical diagnostics, environmental monitoring, and biodefense applications. Like conventional chips that provide speed and cost advantages, biochips have the advantage of using small assay amounts to provide fast detection and lower costs, compared to current assaying methods. An overall top-down design methodology is then proposed for digital-microfluidics-based biochips, with analogues to design methods for conventional chips.
Chapters 2 and 3 present modeling and simulation methods at the "device" level for such systems-specifically, numerical simulation of electrified droplets using droplet electrohydrodynamics, and simulation of electrowetting, which moves, splits, and mixes droplets on a microfluidic biochip. At the other end of the abstraction spectrum, methods for simulation at the system level are discussed in Chapters 5 and 8, using coarser system-level macromodels.
The remaining chapters develop solutions to subproblems within a biochip design flow, employing techniques that recount chip design approaches but that are formulated toward designing biochips. For example, techniques used in on-chip parasitic extraction and simulation appear in several chapters: Chapter 4 describes how to use precorrected FFT methods to simulate the discretized Stokes equation; Chapter 6 presents a boundary-element-based method for solving electrostatics problems in biochips; and Chapter 7 describes the use of model order reduction methods for creating macromodels of bioMEMS components, including those with weak nonlinearities. A neural-network-based modeling approach for bioMEMS is presented in Chapter 9. Concepts from the IC world appear again in Chapters 10 through 12, which develop techniques for floor-planning, placement, and routing. Finally, Chapters 13 and 14 explore reconfigurable systems, and methods for pattern mining using binary decision diagrams, respectively.
Although the two books were apparently independent projects, BioMEMS is as an excellent companion to Design Automation Methods and Tools for Microfluidics-Based Biochips. Whereas the latter primarily gives a CAD perspective in ways that may look familiar to the IC specialist (not surprising, since many of the contributors have worked extensively on IC-related CAD issues), BioMEMS closes the loop with the problem domain by focusing primarily on the design of biochips, with some discussion of modeling issues.
Chapters 1 and 2 of BioMEMS present a history of the field and an overview of microfluidics-based biochips. Although this presentation is complementary to the material presented in Design Automation Methods and Tools, a reader who has already perused that book will find these chapters a comfortable read. Chapters 3 and 4 switch gears to a completely different form of bioMEMS system-neural implants, which insert prosthetic devices that communicate directly with neurons using electrical signaling in case of nerve damage. Electrical models for neural-system communications are presented within this discussion, after which various types of neural prosthetics are detailed.
Chapter 5 provides a deeper look into microfluidic platforms. Likewise, Chapters 6, 8, and 9, respectively, present a comprehensive overview of DNA arrays, protein arrays, and cell-on-a-chip systems for developing cell cultures and functional cell analysis. Chapter 7 details mechanisms and models for capillary electrophoresis, used for separation and detection in LoC systems. Chapter 10 describes arrays of microelectrodes on a chip, for the purpose of cell monitoring in biomedical and pharmaceutical research. Finally, Chapter 11 provides an overview of futuristic applications for biomedical nanoelectromechanical systems (bioNEMS).
Each chapter of this book provides an extensive set of references for further exploration (although the larger number of citations, in comparison with the previous book, may be an artifact of the experimentally intensive nature of this field).
Perhaps the only criticism of Design Automation Methods and Tools is that the organization of chapters sometimes appears not to follow a completely logical sequence. Nevertheless, the reader can certainly come away with a clear big-picture view of the field, as well as an understanding of several details. BioMEMS is coherently organized and is not hard to read: The contributors to this book have successfully abstracted away the biological part, and presented the material in terms that will be familiar to electrical engineers, computer engineers, and computer scientists. The level at which each book operates is fundamentally different: Whereas Design Automation Methods and Tools gets down into the details so that readers have enough information to implement CAD tools, BioMEMS does not attempt to operate at this level in the design domain but instead relies on the extensive citation list. All in all, both books are well written and constitute excellent sources for a first, or second, look at this field.