key: cord-0035477-uw12q19r
authors: Tian, Wei-Chang; Finehout, Erin
title: Microfluidic Diagnostic Systems for the Rapid Detection and Quantification of Pathogens
date: 2008-08-23
journal: Microfluidics for Biological Applications
DOI: 10.1007/978-0-387-09480-9_9
sha: cac9feb28a7902e0a866c833a884fe0782691eea
doc_id: 35477
cord_uid: uw12q19r

nan

pathogens are directly responsible for 15 million deaths worldwide annually. Many of these deaths could be prevented as a result of rapid and/or point-of-care diagnosis. The microfluidic systems reviewed here , which are in use and under development, seek to fulfill this significant need.

The technical objective of these systems is to detect and estimate the concentration of pathogens of interest within clinical samples. This can be achieved by the detection and quantification of any of the following: (a) whole pathogen cells; (b) metabolites released or consumed by the pathogen; and (c) proteins or nucleic acid sequences that are specific to the pathogen of interest. The type of target assayed forms the basis behind the classification of the wide varieties of microfluidic systems encountered. However, irrespective of the actual target assayed for, the microfluidic systems have to overcome twin problems: that of low pathogen concentration and/or the presence of interferents. Hence various strategies, such as culturing, filtration, electro-kinetic separation, magnetic separation, enzyme linked immunosorbent assay (ELISA), polymerase chain reaction (PCR), etc., are used to amplify and/or isolate the target pathogen prior to detection. Detection is accomplished using optical, electrical and mechanical techniques (or a combination of them). In this review, attempts have been made to highlight how the physics at the micro-scale has influenced the design of the separation and detection schemes employed.

Infectious pathogens have posed a considerable threat to global health throughout history. Such pathogens are responsible for the black death (caused by Yersinia pestis) of the fourteenth century and the influenza pandemic that occurred between 1918 and 1920, each of which claimed more than 50 million lives [1] [2] [3] . More recently, the acquired immunodefi- More than 50 new, or newly discovered, pathogens have been identified over the past 30 years [1] . In addition to these new discoveries, there exists the constant threat of re-emerging pathogens, which often occur throughout new or extended geographic regions. West Nile fever is one infectious disease that has recently (1999) made a reappearance within a new geographical location, the United States [4] . This disease is the result of a flavivirus transmitted via mosquitoes. Bioterrorism and the growing phenomenon of antimicrobial resistance further compound the difficulties posed by infectious pathogens. One familiar example of a bioterrorist attack took place in 2001 with the mailing of letters filled with anthrax spores that ultimately claimed the lives of 5 people and caused illness in 13 others [6] . The issue of antibiotic resistance is not necessarily new, but, is growing for a variety of reasons, including the inappropriate utilization of antimicrobials [7, 8] . Infectious pathogens that have developed antimicrobial resistant strains over the last several years include pneumococci, enterococci, staphylococci, Plasmodium falciparum, and Mycobacterium tuberculosis [5] . As can be seen from Figure 9 .1, the prevalence of infectious diseases is a global phenomenon. It is estimated that approximately 25% (15 million of 57 million) of annual worldwide deaths are directly attributable to infectious diseases [2] . The toll of these diseases is much greater in the developing world than it is within the developed world [9] , with at least half of all deaths being caused by infectious pathogens [10] . However, even within developed countries, microbial pathogens are not a trivial public health issue. An appreciation of their significance can be gained by examining relevant sepsis statistics. Sepsis is a condition that occurs when toxinproducing bacteria, and other pathogens, infect the bloodstream. These pathogenic organisms are often introduced into the blood stream via infected organs, such as the kidneys (upper urinary tract infection) and the lungs (bacterial pneumonia), or during a hospital stay via various routes (e.g. intravenous lines, surgical wounds, surgical drains, and open-skin wounds like bedsores). Although it does not attract much publicity, sepsis is one of the leading causes of death in the US, with an estimated 750,000 people annually developing severe sepsis and over 215,000 (close to 30%) of those people succumbing to the infection [11] . Within the US, sepsis claims more lives annually than myocardial infraction [12] , lung cancer, or breast cancer [13] . Perhaps a greater cause of concern is that, unlike most other leading causes of death, the fatality rate associated with sepsis has not decreased significantly over the last few decades [14] . Prevention and early detection leading to prompt treatment are essential since minor infections can rapidly turn life threatening [15] . Multiple studies have shown that the earlier bacteria can be detected within the bloodstream and the appropriate antibiotic can be delivered, the more successful the treatment is in fighting off the infection [16] [17] [18] . Diagnostic systems therefore play an integral role in facilitating an effective response to be provided against these infections.

The ability to detect infectious pathogens is vital to mounting a quick and effective response to outbreaks of pathogen-caused disease. The clinical microbiologist, nurse, or technician is thus often tasked with answering the question, "Are there pathogens present in my sample (be it a clinical sample, food sample, environmental sample, etc), and if so, what is/are their identity/identities, and in what number are they present?" In general, the answer to the first part (what identities) may vary from just a few (< 5) to hundreds, and the number of each type may independently vary over sev-eral orders of magnitude (from a few to millions of colony forming units (CFU) per ml of sample).

The traditional, and most general, technique designed to answer this catchall question, is the plate-culture method [19] . This method was first popularized by Robert Koch in the last quarter of the 19 th Century, and is still extensively used today with only minor modifications. The standard procedure is to dilute the available sample serially over a few orders of magnitude and to spread approximately 0.1 ml of each of these serially diluted sub-samples uniformly over the surface of a petridish (usually 3-4 inches in diameter) containing a bed of agar gel. This gel contains nutrients that support the growth of microorganisms. Once the liquid suspension is spread out, bacteria in the (sub)sample adhere to the agar gel. The plates are then kept at a temperature favoring growth (typically 37ºC), and after an incubation period during which bacteria on the agar proliferate, bacterial colonies that can be visually observed are formed (assuming the initial presence of target bacteria). The number of colonies established is then counted. For a limited range of dilutions of the original sample, it is possible to distinguish individual colonies, and these counts are used to establish the concentration of bacteria in the original sample.

The plate-culture technique is relatively simple and provides a snapshot of the number of bacteria in the original sample at the time that it was collected. However, it does have a number of drawbacks. To begin with, the plating technique utilizes a significant amount of material, is tedious, labor intensive, and takes a long time to provide results (depending on the species and the medium, the time required may range from overnight, to days, or to weeks). The process becomes even more labor intensive if one seeks to screen a sample for multiple bacteria. In this case, multiple petridishes are required and selective media (a medium that supports the growth of only a limited class of bacteria, sometimes even just a single species) have to be designed for each of the target pathogens. Given that in many cases it is not possible to design growth media that are selective for all the individual targets, differences in colony morphology have to be relied upon in order to distinguish different targets and obtain an estimate of their concentrations in the sample. Unfortunately, identifying bacteria based on colony morphology requires very specialized training, and the readings obtained are subject to errors in observer judgment. A more subtle drawback is that the blend of nutrients present in the agar gel may not support the growth and colony formation of all the different types of bacteria present in the (sub)sample spread on the plate. It is also possible that different types of bacteria present may grow/proliferate at very different rates on the agar. It is likely that non-growing or slow-growing bacteria may not be seen, in which case they will not be considered. For many applications, this is a significant limitation. For instance, it is believed that a large fraction of microbes residing in the intestines of humans and other mammals, whose absolute and relative numbers have clinical significance, have never been cultured on plates [20] .

An analogous technique used to determine the viral count of samples is to first grow a "lawn" of bacteria/eukaryotic cells on a petridish, and then spread onto it a suitably serially diluted sample suspected to contain virus particles. As the virus infects and subsequently lyses the cells, clear zones or "plaques" are created in the lawn. Each plaque is analogous to individual colonies of bacteria. However, as is often witnessed with bacterial culture, not only do different viruses vary in their ability to form plaques in the cells used to make lawns, but even different strains of the same virus may also exhibit such differences [21] .

The inherent limitations of culture-based detection methodologies have resulted in interest directed at, and effort devoted to, developing alternate pathogen detection and quantification methods. The main goals of these developed platforms are to speed-up the detection process and to improve accuracy. These aims have resulted in differing technologies that employ tools such as flow cytometry, standard micro-well plate enzyme-linked immunosorbent assay (ELISA), and nucleic acid based techniques. Most diagnostic devices to date, with a few exceptions, have been designed for use within the developed world with well-equipped laboratories and trained technicians [10] . However, there is now a consumer-driven demand for rapid and point-of-care (POC) diagnostic devices, which culture-based and other commonly employed detection techniques will not be able to meet due to long detection times and a lack of portability. Considering the scarce resources that characterize the developing world, such technologies also will not suffice in these regions. In these regions, a detection device must be portable, robust, rapid, and simple to operate.

Microfluidic technologies are ideally suited to meet the challenges described for the rapid and POC detection of pathogens. While the goal of portability is perhaps the leading cause behind the development of microfluidic devices, such systems offer many other advantages. These include reduced detection times, reduced biological sample (and reagent) require-ments, integrated devices that combine sample processing with detection on a single chip, and channel geometries and dimensions that often aid in the isolation and/or detection of the target pathogens.

There exist a large number of microfluidic diagnostic systems that employ different approaches to rapidly establish the presence of the pathogen of interest in a sample of constrained size. In many cases, these techniques also estimate target pathogen concentrations to varying degrees of certainty. In general, these techniques achieve their objective by detecting/measuring any of the following: (a) whole pathogen(s) (b) metabolites released or consumed (c) genetic material or proteins specific to the target pathogen These differing types of targets provide one way of classifying the large number of microfluidic systems either in the market or under development. The next section of this review, which reports on developments in the recent past, has been divided accordingly.

In "real world" samples, the concentration of whole bacterial cells is often low, and/or there are a large number of other interfering species present, complicating pathogen detection and quantification. For instance, at the time when patients begin to manifest clear clinical symptoms of sepsis, they often have bacterial loads of only 1-30 CFU (Colony Forming Units)/ml, and even for patients who are near death, the load may still be as low as ~1000 CFU/ml [22] . In addition, these low numbers of bacteria need to be detected in the presence of blood cells, of which there are millions per milliliter. Microfluidic diagnostic systems thus have to deal with two distinct, but inter-related problems, increasing the concentration of the target pathogen cells (in absolute terms and/or with respect to that of the other cells) and detecting/quantifying their presence. The easiest way to approach this overall problem is to tackle these two challenges individually. First, isolate the target cells and second, devise a platform for detection and quantification. It is expected that during this isolation process, not only will the absolute concentration of targets increase, but that interfering species will be discarded as well.

In many cases, especially those in which isolation of the target species from the sample medium is desired, the specificity of antibody-antigen interactions can be leveraged to enhance the separability of the target species from the sample medium. Antibodies are proteins synthesized by the immune system of animals (including humans) as a targeted response to an infection by a foreign body or pathogen (typically referred to as an antigen). The conformation of these proteins is such that they bind to specific proteins on the surfaces of the invading pathogens. Since different strains of pathogens often exhibit different proteins on their surfaces, it is possible to generate antibodies that are strain specific --such as those developed for Escherichia coli (E. coli) O157:H7, a particularly virulent strain of E. coli [23] . It is also of importance that only a small part of the antibody molecule actually binds to the target antigen. The other end of the molecule, the terminal end, can be chemically modified and immobilized on to a wide variety of surfaces like glass or silica (silicon wafers with a thin oxide layer or glass) [24] , polymers such as acrylates [25] , phthalates [26] , polyethylene [27] , and metals such as gold [28] . As will be discussed later, once target pathogens bind to their corresponding immobilized antibodies, various fluidic, electrokinetic, and other schemes may be used to both quickly isolate them and to rapidly estimate their number.

The following methods (culturing, filtration, magnetic, and electrokinetic) are some of the more common techniques utilized to isolate samples prior to attempts at detecting pathogens present within a given sample.

If the target pathogen is a viable organism, an increase in its number can be facilitated via reproduction by supplying a balanced set of nutrients. Traditional microbiology research has provided standard growth media formulations for wide-ranging bacteria cultures, making it relatively simple to reproduce a bacterial pathogen. There are also special media formulations that selectively support the growth of particular classes of microbes [29] . One such example is Eosin-Methylene Blue (EMB) agar, which is toxic to gram-positive bacteria allowing the growth of only gram-negative bacteria.

Most bacterial cells are slightly heavier than water and other commonly used aqueous growth media, and so the cells will sediment. During aqueous culture, standard practice is then to keep the bacterial cells suspended in the growth medium using shakers, stirrers or other forms of agitation [30] . This also serves to facilitate the transport of oxygen and nutrients to the surface of the cell, and prevents toxic wastes excreted by the cells from building up near their immediate vicinity. Additionally, the cultures are usually maintained at a constant temperature (typically 37ºC) by housing the growth vessels within incubators. Achieving similar local flow regimes and temperature profiles around proliferating cells maintained within reactors at volumes on the order of micro-liters is a non-trivial task. To resolve these challenges, microfluidic valves, mixers, and heaters have to be incorporated to culture the cells [31] [32] [33] [34] . Additionally, sensors to measure and regulate temperature, dissolved oxygen concentration, and pH within the micro-reactor have to be developed and integrated. One such microreactor is shown in Figure 9 .2. Additional challenges in micro-reactor development include considerations relating to the volume, scales, and fabrication materials typically encountered in microfluidic systems. For instance, polydimethylsiloxane (PDMS), a material commonly employed in biomicrofluidic systems, is permeable to air. This material property may lead to sample losses due to evaporation [35] . While the rate may be low (micro-liters per hour or lower), the sample losses cannot be neglected when the operation volume is on the order of micro-liters and the reactor is to be maintained for hours.

Potential corrective measures available include housing the microfluidic device in a humidity controlled box [36] , or specially reformulating the material used to reduce losses due to evaporation [37] . Another issue of concern involves the use of a single micro-reactor for culturing multiple samples, which may be required depending on the cost of the microreactor. Under these circumstances, the formation of biofilms can be problematic. Biofilms are formed when bacteria and other microorganisms adhere to solid surfaces in contact with water and secrete a complex polymeric substance that both anchors them to the solid surface and protects them from bacteriocidal substances in the water [38] . Such biofilms will intermittently shed cells housed within them, which can lead to contamination of the samples being cultured. Biofilms are also difficult to get rid of due to the nature of the polymer matrix in which the microbial cells are housed. The challenges associated with biofilm formation are much more acute in microfluidic systems due to the high surface area to volume ratios encountered [39] , and consequently using disposable systems is preferred. If for some reason a disposable system cannot be employed, complex bioreactor designs may be required to avoid biofilm formation. One such design involves the continuous transport of small, isolated volumes within a micro-channel loop [40] .

As described above, the replication of macro-scale bacterial growth conditions within microfluidic devices is a non-trivial task. Therefore, in designing a diagnostic system that requires pathogen growth through culture, the potential benefits of micro-reactor culture should be weighed against the inherent challenges.

Filtration is conceptually a simple method for increasing the concentration of target species. The sample suspension being investigated is allowed to pass through a filter whose pore size is smaller than that of the target, resulting in the target pathogen being retained in the excluded filter cake (retentate). The retentate may then be resuspended in a volume of liquid much smaller (1/10 th or even less) than the original sample. At the macroscale, filtration is more commonly employed for applications like environmental monitoring and food safety, where large sample volumes (1L or larger) and low bacterial loads (~1 CFU / 100 ml or lower) are commonly encountered. On the other hand, clinical samples are typically much smaller. Given the limited aliquot size that can be introduced into a microfluidic device, the use of filtration at the micro-scale may also be necessitated. For example, consider the case of a liquid clinical sample with a bacterial concentration of 100 CFU/ml. If the aliquot size to be employed within the microfluidic device is 100 μl, then it can be expected that the aliquot will contain 10 CFU on average. However, if the aliquot size is further reduced to 10 μl, then the average colony count within an aliquot will only be 1 CFU, meaning there is a significant chance that the aliquot drawn will contain no bacteria (even though it was present within the original sample suspension). The ability to concentrate the bacteria from the original sample suspension through filtration would thus be beneficial in preventing false negative test results.

Fabricating filters within microfluidic systems poses several challenges, which have been discussed elsewhere [41] . In brief, three types of filters are typically found; arrays of pillars within a micro-channel that are usually arranged in a regular geometric pattern [42, 43] , micron-sized "holes" along one of the walls (usually the "floor") of the channel [44, 45] , and a micro-channel embedded porous matrix that typically provides a distribution of pore sizes [46] [47] [48] [49] .

A single filter with a modal pore size smaller than that of the target pathogen will retain the target, which may then be subsequently assayed [50] . However, if there are too many other non-target species larger than the pore size, this may lead to the system quickly fouling, which complicates subsequent target detection within the filter cake. This issue can be resolved through the on-chip use of two filters, one with a pore size slightly larger than the target pathogen and the other with a pore size slightly smaller than the target pathogen [51] . In this case, the material trapped between the two filters is likely to contain a higher concentration of target particles. It is still possible that a rapid buildup of filter cake on the upstream filter could occur, however, the use of cross-flow and electrokinetic stirring [52] can be employed to remove the cake.

The use of magnetic techniques to isolate biological samples is a familiar technique on the macro scale [53, 54] , and is potentially attractive for use within microfluidic systems for the isolation and manipulation of target pathogens. Of particular interest is the magnetic susceptibility of a particular cell [55] , or of a magnetic particle that can be attached to a particular cell [56] [57] [58] . An example of a cell that exhibits magnetic susceptibility is the red blood cell. Deoxygenated red blood cells are the most paramagnetic cell within the body as a result of the free electrons associated with the iron atom contained within the heme ring of hemoglobin. A continuous separation of red blood cells from whole blood has been achieved in a microfluidic system by placing a small ferromagnetic wire along the length of a microchannel and applying a uniform external magnetic field to create a high gradient magnetic field locally [55] .

In the majority of cases of interest, the target pathogen will not be naturally magnetic. Consequently magnetic particles with the capability of selectively binding to the target pathogen must be employed, and there exist many examples of their use within the literature [56] [57] [58] [59] . Such particles are advantageous for a variety of reasons, including that they offer a considerable surface area for pathogen capture, and that magnetic handling is generally not impacted by changes in pH, temperature, etc. [60] . Magnetic particles are typically composed of an interior magnetic core and an outside shell that can be made of substances ranging from polystyrene to silica to polysaccharide. The idea fundamental to the use of magnetic particles is that some form of interaction between the particle and the pathogen can be created. One prominent example relies upon the conjugation of antibodies specific to a given pathogen onto the surface of the magnetic particle. After the introduction of a sample containing the target pathogen, this pathogen will bind to the conjugated capture antibodies while other substances within the sample should remain unattached. This specific technique was employed by Lien et al in an effort to isolate Dengue virus serotype 2 and enterovirus 7 [58] . This isolation employed Dynabeads ® (Dynal Biotech, Norway), which are hydrophobic superparamagnetic polystyrene beads of approximately 4.5 μm in diameter. After allowing for the target virus pathogen to bind to its respective antibody on the surface of the beads, the beads were then isolated via a magnetic field created using an integrated magnet of microcoils. While holding the beads in place, the remaining suspension could be flushed from the region, allowing for the isolation and concentration of target pathogen.

As described above, the magnetic isolation of a cell or magnetic particle is reliant upon the induction of a magnetic field within the vicinity of the cell or magnetic particle to be trapped. The selection of an appropriate magnet to fulfill this challenge is thus of importance, and in general, either permanent magnets or electromagnets are employed. To date, most work has relied upon supplying a magnetic field from outside the microchannel or microchamber [60] . The development of integrated miniature permanent magnets and electromagnets has progressed and many examples of their use can be found in the literature [59, [61] [62] [63] . However, improving the performance of these internal magnets remains an ongoing endeavor. For a detailed discussion of magnetic theory and magnet integration with microfluidic chips, the reader is directed towards several comprehensive reviews that also provide excellent discussions on the implementation of magnetic nano-and microparticles within lab-on-a-chip designs [60, 64, 65] .

Electrophoresis and dielectrophoresis are the two main available electrokinetic techniques for pathogen isolation. These technologies rely on both the surface charge and the dielectric properties of bacterial cell membranes to isolate the pathogen target from the matrix or suspension in which they are dispersed. Electrophoresis refers to the migration of charged particles suspended in a fluid down the potential gradient of an externally applied electric field. Typically, particles, which are broadly defined here to include both biological cells and macromolecules such as proteins and nucleic acids, migrate with a constant velocity that arises when the viscous drag forces on them equal that of the coulombic forces propelling them down their potential gradients. The velocity of the particle per unit field gradient is referred to as its "electrophoretic mobility". The net charge that the particle carries can be a function of the pH of the solution (most proteins and many cells have a net positive charge at low pH and a net negative charge at high pH), and the pH value at which the particle carries no net charge is referred to as its isoelectric point (pI). Particles with different electrophoretic properties can be separated from each other using various schemes, yielding different techniques such as "field flow fractionation", "zone electrophoresis", "isoelectric focusing", etc. that have been described elsewhere [66, 67] . It has also been known for some time that when suspended in common buffers, bacteria of different species have different electrophoretic mobilities, and that their isoelectric points are also different [68] . These differences have been attributed to variations in the surface composition (surface charge density) between the different species/strains of bacteria [69] . Such differences in the electrophoretic behavior of cells have been extensively used at the macro-scale to separate various types of biological cells from each other and from non-biological material [70] .

Capillary electrophoresis (CE) is a form of electrophoresis where the separands migrate, along with the buffer in which they are suspended, from one end of a capillary to the other due to electro-osmotic flow produced by an applied direct current (DC) voltage. Due to the electrophoretic velocity of the seperands relative to the buffer, the seperands elute at a dif-ferent time than when they would have had they flowed along with the buffer. By definition, CE is a "microfluidic" technique. Though traditionally performed within specially constructed capillaries, it can also be performed within channels housed in microfluidic chips [71] . The narrow dimension of the capillary (or micro-channel) plays a key role in the success of the technique by preventing the development of recirculation vortices that tend to arise due to temperature differences generated by the joule heating of the ionic current through the system. The Grashof Number, which gives the ratio of buoyancy forces to viscous forces on a fluid element, predicts the formation of natural convection vortices, and scales as the cube of the characteristic length. Consequently, reducing the characteristic length of the system suppresses natural convection that is driven by temperature induced density differences. The high surface area to volume ratio within a microfluidic device also enables relatively rapid heat transfer between the capillary interior and the ambient, thereby further preventing the onset of natural convection.

CE has been used extensively to sort "bioparticles", such as viruses, bacteria, eukaryotic cells, and even sub-cellular organelles. Detailed reviews are available that summarize the results obtained by various researchers in this area [72, 73] . Due to their varied electrophoretic mobilities, different types of bacterial cells elute at different times from the capillary. The species being eluted can be detected using methods such as UV absorbance or laser induced fluorescence [74] . Capillary zone electrophoresis (CZE), which is the most commonly employed form of CE, is characterized by the maintenance of a constant pH along the whole separation pathway. It has been used to separate significantly differing bioparticle types, such as gram positive and gram negative bacteria [75, 76] . A more sophisticated form of CE, known as capillary isoelectric focusing (CIEF), establishes a pH gradient along the separation path and has been employed in the more difficult separation of bacteria of similar sizes [77] .

There are drawbacks, such as lowered throughput, to the use of CE. Another significant disadvantage with CE stems from the fact that extremely high voltages (kilo-Volts) must be used to drive the electro-osmotic flow (EOF) through the capillaries. This high voltage requirement not only necessitates the use of bulky (and dangerous) high voltage supply units, but also produces extensive electrolysis (gas bubbles). Special provisions must then be included within the device to allow for the escape of air [78] , while also compensating for the fact that if air bubbles do enter a micro-channel or capillary, then EOF alone may not be sufficient for displacing the bubbles within the conduit [79] .

To avoid the drawbacks inherent in using DC methods such as electrophoresis, a method like dielectrophoresis (DEP), that requires the use of alternating current (AC) rather than DC voltages can be employed. In contrast to DC, within AC the polarity of the electrodes changes once every halfcycle of the AC signal. If frequencies higher than ~10 kHz are used, the half-cycle time period is shorter than the charge transfer time at the electrode for most electro-chemical reactions, and consequently most Faradaic reactions, which produce air bubbles, are suppressed [80, 81] . In characterizing DEP, it is convenient to consider it the AC analog of electrophoresis. DEP is a process in which migration of suspended particles (not necessarily charged) occurs in the presence of an electric field gradient, either towards or away from regions of high electric field strength. The migration of a particle towards a high-field region is termed positive DEP, while the opposite is known as negative DEP. Whether the particle experiences positive or negative DEP is determined by the sign of the Claussius-Mossotti factor, a complex term that depends on the values of the conductivity and dielectric constants (permittivity) of the particle and the suspending medium [82] .

At low frequencies (ω<1 KHz), a particle's dielectrophoretic behavior is almost completely dominated by differences in conductivity, while at high frequencies (ω>100 MHz) its DEP behavior is dictated by differences in permittivity. However, for frequencies that can be practically realized, ~10 kHz to 100 MHz (high enough not to generate bubbles, but low enough to not require expensive equipment), DEP behavior depends in a complex manner on both of these quantities. Qualitatively though, the effect of conductivity is more pronounced at lower frequencies and that of permittivity at higher frequencies. In standard buffers, such as phosphate buffered saline (PBS), both the conductivity and the permittivity of bacterial cells are typically lower than that of the medium. Consequently, bacterial cells generally do not display any cross-over from negative to positive DEP (have negative DEP for all frequencies). However, when these cells are placed in low conductivity suspensions (~100 mS/cm), they display positive DEP at lower frequencies and cross-over to negative DEP in the range of 100 KHz to 100 MHz [83] . Since dielectrophoretic phenomena rely on field-induced dielectric polarization at the cell membrane, and the composition and properties of the membrane vary from species to species, the cross-over frequency (the particular frequency at which the DEP behavior changes from positive to negative) is generally a function of the species and strain of the cell [84] . Thus, there can exist certain frequencies where the DEP behavior of target cells is different from that of other cells (and non cellu-lar impurities in the system), allowing for the isolation of the target cells from complex suspensions containing many other species [85] [86] [87] [88] . However, it must be considered that prolonged storage, or other sample handling processes that impact the physical composition of the membrane can change the cross-over frequency of the target cell [89, 90] .

Perhaps the main disadvantage associated with DEP is that typical obtainable particle velocities are low. It can be shown that the DEP force on a particle scales to the third power of the particle radius and quadratically with the applied voltage. This cubic dependence on particle radius in conjunction with practical limitations on the applied voltage render the particle velocity produced by a DEP force miniscule, typically on the order of 10 μm/sec for bacteria and 1 μm/sec for viruses. Concentration times on the orders of hours are thus required. Also problematic is the fact that the field gradient necessary to drive DEP motion can only be achieved with relatively narrow inter-digitated electrodes, whose field penetration depth is limited by the electrode width. As a result, DEP channels are usually less than 50 microns in transverse dimension. The slow capture and the small transverse dimension produce extremely low throughputs for continuous flow kits. For example, for a 5 cm-long device employing a 50 micron transverse dimension and a 10 micron/s DEP velocity, pathogen capture can only be ensured if the throughput is less than 25 nanoliters per second. At this flowrate, it would require more than 10 hours to process a 1 ml sample. This low throughput and batch volume can be somewhat alleviated with a massively parallel array for laboratory use, but such arrays cannot be easily designed for a miniature and portable diagnostic kit.

Bacteria isolation and concentration by DEP can be sped-up by using fluidic forces (convection) to transport suspended particles (of all types, including the target pathogen) to a specific localized region, and then capturing the targets utilizing DEP forces. This basic idea has been extensively applied in a variety of manners to capture different targets by the corresponding author of this review and collaborators. This can be accomplished by generating converging flows with stagnation points, and imposing the electric field gradient on the particles at the electrode stagnation points (or lines) where the viscous drag is weakest. In such systems, while all suspended particles are swept towards the stagnation point along with the fluid flow, only particles (such as the target pathogen) that are directed by the electric field (due to positive or negative DEP) towards the stagnation points are captured. Converging flows of this nature have been generated by AC EOF at symmetric electrodes [91] , by AC EOF generated with serpentine coils [92] , by rotational vortices generated using an ionic wind [93] , and by a combination of DC bias and circulatory AC EOF [94] . It is possible to generate fluidic velocities of approximately 10 mm/s, which represents a thousand-fold increase in the speed at which target particles move towards the collection zone compared to simple DEP. Once collected at a "point" (a volume of a few pico-liters), the pathogens can then be detected via a variety of detection mechanisms. The rapid dielectrophoretic isolation and/or concentration of pathogens within continuous flow-through systems has been accomplished utilizing methods such as integrated "3-dimensional" (3D) electrodes. Such designs employ aligned electrode pairs fabricated either along the sidewalls [96] or on the top and bottom [97, 98] of channels through which the particleladen fluid flows. This 3D configuration produces a high field normal to the flow direction and penetrates across the entire height of the channel. Therefore, bacteria present through the entire depth of the channel (as opposed to just those near the bottom, as is the case for single electrode devices) can be deflected and/or captured while liquid flows unimpeded. Through tailoring the orientation of the 3D electrode, the particles can be directed to different streamlines/exits, or can be prevented from flowing past a point in the channel altogether, creating a DEP force field cage through which particles can not pass [99] [100] [101] [102] . DEP fields from 3D electrodes have been exploited to generate a continuous sorting and trapping device, which is shown in Figure 9 .3, that is capable of sorting and collecting three different types of pathogens at a rate of ~300 particles/sec [95] . The unique physical and electrical properties of nano-particles, such as carbon nano-tubes (CNTs), can also be harnessed to isolate target species dispersed in the sample. Due to their ability to store charge and their long, narrow geometry, very strong dipoles are induced by exposing CNTs to a high frequency AC field. This focused field, especially at the tips of the CNTs, leads to high field gradients, which serve to attract pathogen particles to the CNTs by DEP. CNTs may be embedded onto the "floor" of the micro-channel [104] , or may be deliberately dispersed into the pathogen containing suspension [103] . In the latter case, on average the CNTs are closer to the target pathogens. Additionally, at lower AC frequencies, CNTs are themselves attracted to electrodes embedded within the microfluidic device. Their cylindrical geometry confers upon them a much higher DEP mobility than is typically observed for spherical or ellipsoidal pathogens. Consequently, suspension dispersed CNTs will drag along the surface of embedded electrodes as they flow through a micro-channel, thereby bringing trapped pathogens to the electrode surface. This approach, as illustrated in Figure 9 .4, has been successfully utilized to enhance the trapping rate of target pathogens by a factor estimated to be over 1000.

In an aqueous solution, charge is carried between any two electrodes by ions in the solution. The presence of particulate matter (such as pathogens) physically obstructs the movement of these charge carrying ions, and thereby leads to a higher resistance (or impedance, when using an AC signal) between the electrodes. Typical pathogen loads (~1000 particles/ml) have a volume fraction of ~10 -12 , and consequently this change in resistance/impedance is not significant. It may, however, be discernable if the suspension is made to pass through a narrow slit, only slightly larger than the pathogens, that possesses electrodes on two opposite ends [105] , or if the target pathogens can be made to adhere to or congregate at a surface on or very near an electrode. Quite often, the surface of interest may be treated to enhance the adhesion of cells [106] , or specific antibodies may be immobilized to capture (and hence detect) only the target pathogenic entities [107, 108] .

The above mentioned electrical detection techniques suffer from drawbacks that make it difficult for them to handle "real world" samples. The former, passing the suspension through a narrow slit with electrodes on two opposite ends, is plagued by the high likelihood of the slit being clogged by random particulate matters present in field samples. The latter method, forcing the target pathogens to adhere or congregate at a surface on or near an electrode, suffers from two main drawbacks. First, particles other than the target may get deposited or adhere non-specifically to the surfaces being monitored, thereby yielding misleading results. Second, the targets may take a long time to diffuse to the surface, thereby degrading the performance of the system. This latter problem can be overcome by coupling the detection system to the systems described above for isolating the targets [109] .

Another potential approach is to use a specific property of the pathogens that enable them to provide a unique and recognizable signature. One such property that can be exploited is the ability of bacteria to store charge. Direct measurements [110] of the amount of charge in an individual bacte-rium (~10 -13 Coulombs) and their zeta potential (~20 mV) suggest that an individual bacterium, if considered a solid sphere of radius 1 μm, has an effective dielectric constant of ~10,000. This remarkably high number probably arises due to the presence of proteins (such as those forming the various ion channels and regulators) that are extremely efficient in capturing electrons. In contrast, the dielectric constant of aqueous solutions is typically about 80, and so the presence of even a few bacteria should greatly alter the dielectric constant of the solution. Unfortunately, the "double layer" capacitance at the electrode solution interface effectively screens this bacterial capacitance that is dispersed in the bulk solution, especially when operating at frequencies below 1 MHz [111] . However, by modulating the geometry and positioning of the electrodes in a manner that increases the bulk resistance (R), and hence increases the RC time of the medium, the measured reactance ("imaginary" or "out-of-phase" component of the measured impedance) can be made sensitive to bulk capacitance. Consequently signatures of bacterial presence and proliferation at concentrations low enough (100-1000 CFU/ml) to be applicable to many "real world" problems have been detected [112] .

Optical detection strategies are arguably the most widely employed platforms for microfluidic analyte detection. Such strategies are diverse, and examples of pathogen microfluidic detection based on absorbance [113] , chemiluminescence [114] , fluorescence [115, 116] , light scattering [117] , Raman spectroscopy [118] , and refractive index techniques [119] can be found in the literature. A comprehensive treatment of these techniques is beyond the scope of this text, and consequently those readers looking for a rigorous discussion are referred to many excellent reviews [120] [121] [122] [123] [124] [125] [126] .

Each of the detection techniques listed above have specific advantages, however, fluorescence detection schemes are the most widely employed. Such techniques are commonly utilized in conjunction with microfluidic immunoassays [127, 128] , electrophoresis separation schemes [116, 129] , and polymerase chain reaction (PCR) amplification of DNA [130, 131] . Fluorescence is popular due to its high sensitivity, allowing for single molecule detection [132, 133] , high selectivity, and easy integration of fluorescent tags within the microfluidic system. The joining of fluorescence-based detection platforms with on-chip assay and separation schemes has become prevalent. For instance, Xiang et al. produced an immunoassay device that was capable of detecting E.coli O157:H7 at a limit of detection of 0.3 ng/μl [127] . A membrane-based immunoassay system was developed by Floriano et al. that incorporated fluorescent detection of Bacillus globigii at a detection limit of as little as 500 spores [128] . Realtime PCR chips that employ fluorescent detection schemes have been developed for the amplification of hepatitis B virus [130] and E. coli O157:H7 stx1 [131] . Electrophoretic separation of PCR amplicons and subsequent fluorescent identification has been demonstrated for the detection of E.coli O157:H7 and Salmonella typhimurium at a starting concentration of as low as 6 copies of target DNA [116] .

While fluorescent-based detection schemes are popular within microfluidic diagnostics, the majority of these techniques still rely on relatively bulky external equipment, such as confocal microscopes, preventing them from truly meeting the goals established for an ideal microfluidic diagnostic device. Most commonly, a microscope, connected to either a CCD camera or a photomultiplier tube (PMT), is focused onto the microchip at an appropriate location [125] . Efforts to miniaturize optical detection platforms are currently underway, and are discussed in detail elsewhere [120, 126] . Light emitting diodes (LEDs) coupled with optical filters and silicon photodiodes offer the potential for developing a miniaturized optical detection platform [134] . In producing a hand-held device designed for real-time PCR, Higgins et al developed a miniature optical platform employing two LEDs, at wavelengths of 490 nm and 525 nm, along with a miniaturized PMT [135] . This system rapidly and successfully detected Bacillus anthracis and Erwinia herbicola.

Flow cytometry is an integrated technique for counting, examining, and sorting particles suspended in a fluid. There are three key components of a typical flow cytometer, the first being a fluidic mechanism that causes all of the particles in a suspension to line up in a single file as they flow down a channel. The second key component is a set of detectors (such as lasers of different wavelengths) that can probe individual particles, along with the fluid stream, flowing past the detector and obtain information (such as whether a given cell has taken up a particular fluorescent dye) that indicates one or more specific properties of the cell. This information, along with the known velocity of the particle traveling down the channel, can then be taken advantage of by the third key component of the flow cytometer to steer target particles/cells to specific downstream collection chamber(s), while the remaining particles are discarded. Although conventional state-of the-art flow cytometers can measure and subsequently sort particles based on a combination of as many as ten parameters [136] and/or achieve throughputs as high as ~10,000 cells per second [137] , they do suffer from a number of drawbacks. Besides requiring expert operators, they also require large volumes of sheath fluid (~1 L of sheath fluid per 1 ml of sample) and high performance pumping systems to operate, thereby making them non-portable and prohibitively expensive for routine diagnostic procedures in the clinical setting.

Extensive research has been conducted to design and fabricate miniature versions of flow cytometers that replace conventional glass capillary based systems with microfluidic chips that employ integrated optics and hydrodynamic or electrokinetic based flow-switching systems for collecting cells of interest. An excellent review of microfluidic flow cytometry has been provided by Huh et al. [138] . These research efforts, along with others directed towards developing microfluidic pumps, valves and other peripherals [139] [140] [141] , have already yielded commercial bench-top flow cytometers, such as the Agilent 2100 Bioanalyzer® (Agilent Inc.), the Cyflow® ML (Partec GmbH) and the Mycrocytometer® System (Micronics Inc.). While these systems, like their macroscale counterparts, are primarily designed for mammalian cells, they have also been employed by some researchers to quantify bacteria [142, 143] . This is a difficult task though as bacterial cells are typically 10-100 times smaller than mammalian cells, making them difficult to confine to a single file in a fluid stream, and thus providing unsatisfactory results. Consequently, there have also been efforts to develop cytometers specifically for bacterial cells [143] [144] [145] .

The main drawback of most of these portable microfluidic cytometers (including those under development) is that they typically handle very small volumes of sample (for instance, 10 μl for the 2100 Bioanalyzer), which may not be suitable for operating a real-world sample. Research to increase the throughput of such devices continues, especially for bacterial diagnostics [146] . As is the case for other microfluidic flow-through systems, the likelihood of particulate contaminants in real word samples clogging the device, or otherwise adversely affecting measurements, remains a challenging problem.

Cantilevers are beams that are fixed at one end and free at the other. The weight of the beam (when subjected to gravitational forces) and external forces, if any, flexes the beam. The beam's elasticity, in turn, attempts to restore it to its original shape, thus leading the system to have a characteristic natural frequency of vibration. This frequency, known as the principle resonant frequency, is a function of the effective mass and the spring constant of the beam. Therefore, when the effective mass of the beam in-creases due to the adhesion of particles (like bacteria) to it, the vibration of the beam is dampened, and the principle resonant frequency decreases [147] . This effect is utilized to detect the presence of bacteria of interest in many microfluidic systems. Typically, antibodies specific to the target pathogen are immobilized on the cantilever, either by physical adsorption [148] or chemically [149] , prior to exposing the system to the sample of interest. Ideally, washing with plain buffer removes all particles adhering non-specifically to the cantilever, and the resonant frequency can subsequently be measured. This is typically accomplished through an external instrument, such as an "optical lever" or interferometer, or through changes in the piezoelectric resistance or capacitance of the beam as measured via internally embedded circuit elements [150] . The change in the resonant frequency from its value prior to the introduction of the sample of interest is related to the amount (mass) of bacteria adhering to the cantilever, thereby also allowing for a quick quantification of the target species as well.

Cantilevers can be fabricated in silicon based systems using either bulk [151] or surface [152] micromachining. These techniques allow the fabrication of multiple cantilevers within a small surface area [153] . Since different cantilevers can be coated with different antibodies [154] , the potential for simultaneous detection of multiple pathogens exists. However, coating multiple antibodies on adjacent micro-cantilevers is a complex task, requiring multiple surface treatment steps (one for each type of antibody being coated/immobilized). Furthermore, this process is made more difficult by problems of "stiction" (fluid remaining stuck underneath the cantilever due to surface tension), that require special treatment procedures like freeze-drying [155] , "critical point drying" [156] , application of laser pulses [157] , etc. after each step involving a different solution.

Once these systems are deployed for biosensing, the cantilevers are subjected to multiple stresses that affect their performance. Flow and mixing of the sample solution can give rise to shear forces that bend (and may even break) them. Non-specific adsorption of random species from the sample and electrochemical reactions can change the mass of the cantilever. Moreover, only some of these effects may be corrected for using a reference (control) cantilever in addition to the sensing cantilever. In addition to sharing the problems associated with non-specific adhesion with most antibody-based systems, having their antibodies anchored to a nondispersed surface also requires that the targets have to diffuse to the surface (thereby requiring a long time with solutions having low counts of the target species). Thus, while they have been used in the laboratory to detect bacteria suspended at reasonably high concentrations (10 6 / ml and above) in buffers [148, 151] , such cantilever based systems may not have the degree of ruggedness that will be required to handle a large number of real world samples.

When living pathogens respire, they take in sugars, proteins, and other molecules from the environment and release carbon dioxide, pyruvic acid, and large numbers of other compounds. This changes the properties, like pH, dissolved concentration of gasses (oxygen and carbon-dioxide), and electrical conductivity, of the suspending medium. If an appreciable change is detected in these properties, it is taken to imply that pathogens are present that are causing the observed change. A generic growth medium (such as Tryptic Soy Broth, Luria Broth, or Beef Broth) supports the growth of multiple pathogenic and non-pathogenic microbes. Therefore, the use of such media only allows the determination that there exist some viable cells within the sample of interest. That being said, however; selective media can be designed that supports the growth of a limited number of pathogens, typically being those from the same genus. For instance, the Cornell Modified Growth Medium supports the growth of Mycobacteria (a slow growing class of bacteria whose members include the organism causing tuberculosis in humans) while suppressing the growth of other microorganisms [158] . The elapsed time prior to detection (of a specified change in medium properties) can be used to predict the initial pathogen load. This is fairly intuitive given that the greater the initial number of microorganisms in the sample, the higher will be the overall rate of metabolism occurring within the sample, and the faster a detectable change in medium properties will occur. The initial load against elapsed time to detection (TTD) calibration for a particular system (device-medium-organism) can be obtained relatively easily. These methods span a variety of techniques, from those based on detecting the activity of a specific enzyme [159] , the release of a specific metabolite like carbon dioxide (radio-labeled [160] or otherwise [161] ), looking for an increase in the electrical conductivity [162] , a change in the pH (typically observable as a color change within most culturing medium) [163] , or finally a change in the oxygen tension of the medium as a whole [164] .

A large number of high throughput, but not necessarily "microfluidic" devices rely on this basic principle. These include the Bactec,™ which de-tects the amount of radio-labeled carbon dioxide released; the ESP® Culture System (TREK Diagnostic Systems), which detects a decrease in oxygen tension; Coli-Check TM swabs, which use Bromocresol Purple as an indicator to measure a decrease in pH due to bacterial metabolism, and the Bactometer™ (Bactomatic Ltd.), Malthus 2000™ (Malthus Instruments Ltd.), and RABIT™ (Don Whitley Scientific Ltd.) systems that each use electrical impedance. These devices are typically designed to handle samples of ~1-10 ml, and are used for a variety of applications, such as environmental water quality testing, food safety, and veterinary and medical diagnostics. As these devices/methods have greatly simplified handling procedures and incorporated many other features (such as automated readouts) that aid in enabling high-throughout operation, they have acquired broad acceptance in the applied microbiology community. However, the common operating principle, monitoring the sample for effects of microbial metabolism, fundamentally limits how quickly these devices can deliver reliable results. In other words, their "time-to-detection" is still quite high, often being comparable to that of the traditional plating technique. In some cases, such as when the initial bacterial load is very low (~100 CFU/ml or lower), the automated techniques may actually take more time than the traditional plating technique (in which the TTD is independent of bacterial load). This is because the amount of metabolite processed by an individual bacterium is extremely small. Given that the specific oxygen consumption rate for E. coli has been estimated to be ~20 milli-moles of oxygen per hour per gram (dry weight) of bacteria [165] , and a typical bacterium has a dry weight of ~10 -12 grams [166] , a fairly active individual E. coli bacterium consumes ~2 x 10 -14 moles of oxygen per hour. A typical well-oxygenated sample will have a dissolved oxygen concentration of ~2 x 10 -4 M (2 x 10 -7 moles in 1 ml of sample). If the sample has a moderately low initial concentration of bacteria to begin with (~1000 CFU/ml), then the hourly consumption of oxygen for 1000 bacteria (~2 x 10 -11 moles) represents a 0.01% change in the concentration of dissolved oxygen. Such a change is so small that it is practically undetectable. Given the exponential growth nature of bacteria, a subsequent change in the concentration of the metabolite being monitored of 2-5% in 8-12 hours for E. coli would be expected. At that point, the presence of the original 1000 could be confirmed. However, in this case, the TTD would be comparable to that obtained with the traditional plating technique. Lower initial loads would further increase this TTD, and slower metabolizing/growing bacteria would further stretch the TTD. Consider that the detection times for mycobacteria are typically 10-45 days using such automated techniques [164] .

Many microfluidic counterparts of these above devices essentially apply the same principles to micro-liter (or smaller) volumes rather than the 1-10ml volumes characteristic of the above devices. Bashir and co-workers have used both impedance [167] and pH [168] as the quantities monitored to register the occurrence of bacterial metabolism. Gas sensors [169] and biosensors for specific chemicals [170] have also been incorporated within microfluidic systems for detecting/monitoring bacterial growth and proliferation. Interestingly, when cells are confined to small volumes (on the order of picoliters) and consequently are in close proximity to the electronic sensing elements, it has also been possible to use thermal sensors [171] to record bacterial metabolism. The TTD of the microfluidic versions of the metabolism based detection systems, however, is limited by the same basic principle that stymies their macrofluidic counterparts.

As discussed earlier, if large aliquots need to be sampled due to concerns arising from expectedly low bacterial loads, a pre-concentration step may have to be performed upstream of the reactor. Devices have been developed [86, 172] that employ dielectrophoresis to concentrate bacterial cells from a comparatively larger volume to a smaller one where they are subsequently incubated. In general, as shown in Figure 9 .5, there exists an inverse logarithmic relationship between initial bacterial load and time to detection (TTD). This concentration process cuts the TTD by increasing the initial load.

Despite the benefits provided by pre-concentration, monitoring bacterial metabolism may not be the most sensitive way to detect the presence of pathogens as evidenced by the high TTDs that are typically encountered. However, this detection platform does have a number of other redeeming features, such as the relative ease with which these devices may be fabricated, assembled, and coupled to automated fluid handling systems and electronic readouts. In addition, there exists the advantage of the recognition within the diagnostic community at large of principles involved within these devices that have been previously utilized in pathogen detection strategies. The Association of Official Analytic Chemists International (AOAC) subjects new techniques to two stages of approval: a "first action" that allows its incorporation in commercial devices under supervision, and a "final action" that approves the method for widespread use after the new method has satisfactorily met its benchmarks in the commercial setting. Impedance based methods for detecting bacterial metabolism in microfluidic systems have been accepted as a "first action" method. It would hence not be surprising to see a commercial microfluidic device of this type in the near future. 

Immunoassays and nucleic acid based assays are two vital pathogen detection technologies, and consequently efforts to create miniaturized counterparts are significant. Immunoassays, protein binding assays utilized for the detection of antigens and antibodies, are most commonly carried out as an ELISA. Nucleic acid based assays utilize genetic material, most commonly DNA (and sometimes RNA), for detection and quantification. The fundamental methods developed for nucleic acid diagnostics are enzymatic DNA restriction, nucleic acid hybridization, PCR, and fluorescence-based techniques [173] . Of these techniques, DNA amplification procedures utilizing PCR and PCR-based methods (real-time PCR, strand displacement amplification, nested PCR, etc) are of particular interest given that DNA is present at a low concentration within a typical biological sample. Conse-quently the genetic identification techniques discussed within this section will focus on microfluidic PCR-based technologies. Several excellent reviews of pathogen identification methods through diverse nucleic acid based assays can be found [173, 174] .

Immunoassays are arguably the most heavily utilized quantitative pathogen detection platform [175, 176] . These assays take advantage of the specificity of formation of an antibody-antigen complex, which is a result of the fact that such antibodies are produced based on a specific antigen. Therefore, in use within a diagnostic assay, a given antibody will remain specific to the antigen for which it was developed. As mentioned above, one of the most fundamental immunoassay techniques is the ELISA. There are several different ELISA types, including direct, competitive, and sandwich formats. The sandwich ELISA will likely be one of the primary immunoassay platforms employed on-chip given that it generally exhibits impressive sensitivity and specificity traits, along with favorable kinetics [176] . This assay, as it is traditionally performed, is presented in Figure  9 .6. Basic schematic (minus the rinsing and blocking steps) for a traditional sandwich ELISA. In the first step (A), a capture antibody specific to the target antigen is attached to the surface of a microwell plate. This is followed with a rinsing step, a blocking step, and an additional rinsing step. Sample solution, containing target antigen, is added next, allowing for the target antigen to bind to the capture antibody (B). After an additional rinsing step, a detection antibody, which is generally covalently bound to an enzyme, is introduced and allowed to selectively bind to the captured target antigen (C). Finally, after a last rinse, a substrate molecule is added and converted into a product molecule via the bound enzyme (D). This product molecule is utilized as a means of detecting and quantifying the target antigen.

ELISA assays have become commonplace within diagnostics and have been tailored to a wide variety of detection targets. These assays are designed to detect an antigen directly associated with the pathogen or an an-tigen/antibody produced within a host system as a result of the invasion of the pathogen. Several examples include the detection of Escherichia coli (E. coli) O157:H7 [113, 177, 178] , exotoxins from Salmonella and Clostridium botulinum [179, 180] , and Helicobacter pylori [117] . One commercial detection platform based on a sandwich ELISA assay is the VIDAS system marketed by Biomerieux Inc. [178] . This system, which captures bacteria utilizing an antibody specific to an antigen expressed on the surface of the cell, is capable of detecting campylobacter, E. coli O157:H7, Listeria spp, Listeria monocytogenes, Salmonella, Shigella, and Staphylococcal enterotoxins [178] . Another commercial use of the ELISA is within HIV detection. Currently, the majority of HIV testing centers within the United States employ an ELISA that targets antibodies produced as a result of the invasion of the virus [181] . The sensitivity and specificity of this test are both greater than 99% [181] . A positive ELISA result for HIV is typically followed up with a repeat ELISA and then a Western Blot test for final validation.

Conventional ELISA assays are typically performed on a multi-well p and are static. Hence the transport of the target molecule to the reactive surface is through diffusion. This process is hampered in terms of being both labor and time intensive, with assay times ranging from at least a few hours to a few days. These drawbacks can often lead to inconsistencies in the obtained assay results and are problematic in producing a rapid diagnostic assay that can be utilized for POC healthcare solutions. The development of microfluidic ELISA platforms will likely relieve these stresses. While microfluidic-based ELISA has not received the same attention that nucleic acid based microfluidic technologies have, the trend is beginning to change [182, 183] . Commercial microfluidic immunoassays are beginning to appear, with one example being the Triage® system developed by Biosite® Inc. (San Diego, CA, USA) to detect several targets, including Clostridium difficile. Microfluidic devices are attractive platforms for immunoassays given that they can alleviate the time and labor difficulties of conventional ELISA as transport is through both diffusion and convection, transport distances are greatly reduced, and the assay can largely (if not completely) be automated. Moreover, such devices have the potential to be fabricated and operated inexpensively, and provide the opportunity to integrate the assay with on-chip sample preparation and target detection, providing a micro-Total Analysis System (μ-TAS).

The majority of developed microfluidic-based ELISA platforms have either utilized the surface of a microchannel [113, 176, [184] [185] [186] or the surface of numerous micro-or nano-beads that are trapped by various means within a microchannel or a microchamber [187] [188] [189] . One of the key issues involved in localizing an immunoassay to the surface of a microchannel wall is the efficient immobilization of antibody (or protein) to the surface while preventing non-specific protein adsorption that would both interfere with the assay and also reduce the available surface area for target protein immobilization. This issue was discussed in a review written by Lee [190] and has been considered in several publications [176, 185, 191] . In developing a microfluidic immunoassay device for the detection of E. coli O157:H7, Bai et al. functionalized a poly(methyl methacrylate) (PMMA) surface with poly(ethyleneimine) (PEI) to introduce amine groups to the polymer channel walls [191] . This was followed by glutaraldehyde addition to provide aldehyde groups for protein binding. An affinity purified antibody to E. coli O157:H7 cells was then allowed to adsorb to the surface of the channel via the aldehyde groups, followed by the introduction of a sample suspension containing inactivated cells which was allowed to bind to the immobilized antibody. Finally, peroxidase-labeled affinity purified antibody to the cells was added. The peroxidase enzyme catalyzed a substrate to product conversion, producing fluorescence for detection purposes. This surface treatment method improved antibody binding, and compared to untreated PMMA microchannels, an approximately 45 times greater signal and 3 times greater signal/noise ratio was achieved. Additionally, this microchannel device required only 2 minutes for E. coli O157:H7 capture and required a mere 5-8 cells within the initial sample aliquot, both of which are an improvement compared to traditional multiwell plate ELISA.

The use of beads within an immunoassay microfluidic device increases the complexity of the device and requires methods for trapping and handling the beads on-chip [192] . However, the use of beads is becoming more commonplace [187] [188] [189] 193] because of inherent advantages, such as an increased available surface area to volume ratio (which in-turn increases the real-estate that can be taken advantage of for protein binding), further reduced antigen-antibody molecular diffusion distance, potential for multiplexing, ease of manipulation, and ease of performing surface chemistry modifications. One specific example of a bead-based microfluidic immunoassay platform was provided by Liu et al. for the detection of a marine iridovirus [187] . In their work, antibody-coated microbeads (~2.89 μm in diameter) were trapped within a microfluidic channel via an included filter assembly. After incubation with the iridovirus, detection was provided through the insertion of quantum dots attached to the detection antibody. In comparison to a traditional ELISA scheme, this approach was able to detect the marine iridovirus at a significantly improved limit of detection (22 ng/ml as opposed to 360 ng/ml), in a shorter time period (<0.5 hours as opposed to >3.25 hours), and with less demand on the availability of antibodies (0.035 μg used as opposed to 0.5 μg).

The identification of pathogens through genetic analysis techniques is becoming more conventional. Several examples include the detection of Helicobacter pylori [194] , Mycobacterium tuberculosis [195] [196] [197] , Neisseria gonnorhoeae [198] , Streptococci (Group B) [199] , and E. coli O157:H7 [200] . The genetic identification of pathogens is dependent upon the presence of sufficient genetic material (i.e. DNA or RNA). However, it is typical that the target pathogen is present at a dilute concentration, complicating detection. To alleviate this difficulty, DNA amplification becomes essential. This is typically achieved through implementation of the PCR, which incorporates three cycled reaction steps: a denaturing step run at 92-96ºC, an annealing step run at about 40-65ºC, and finally an extension step run at about 72ºC.

PCR-based methods for pathogen detection have gained popularity over traditional culturing procedures, primarily due to enhanced assay speed and reliability improvements. The same push for improved assay speed and reliability are now driving the development of microfluidic PCR-based pathogen detection technologies, with the ultimate goal being the development of POC diagnostic devices that meet the goals laid out earlier.

Several of the advantages offered in miniaturizing the PCR process are similar to those expected with the miniaturization of immunoassays and include a reduction in sample and reagent volumes and a corresponding reduction in reagent costs, increased portability and disposability, and integration with upstream sample preparation and downstream analysis techniques to provide the highly desired μ-TAS. One notable advantage specific to microfluidic-based PCR is a decreased time of amplification as a result of low thermal capacities and large heat transfer rates (because of large surface area to volume ratios) [201, 202] . Though the miniaturization of PCR offers several advantages, the first micro-PCR device was not described until 1993 [203] . Currently, microfluidic chip based PCR can operate with reaction volumes of less than 200 nl [204] [205] [206] [207] [208] , and can offer reductions in reaction time through improvements in the on-chip heating and cooling rates. A conventional thermal cycler provides heating and cooling rates of approximately 1-2ºC/s, however, a heating rate of 175ºC/s and a cooling rate of 125ºC/s has now been reported within a microfluidic design [209] . Commercial microfluidic based PCR devices have arisen as well, and have been employed in detecting a variety of pathogens. For instance, Mycobacterium bovis was detected using the GeneXpert ® system produced by Cepheid (Sunnyvale, CA) [210] . Within this system, a filter is employed to capture organisms from a clinical sample, which is then followed by cell lysis. Real-time PCR of the released nucleic acids can then be conducted for pathogen identification. This system is of particular interest as all of these steps, organism isolation, cell lysis, and real-time PCR detection are performed on a single, disposable cartridge. Another example of a commercial microfluidic based PCR device is the BioMark™ 48.48 Dynamic Array developed by the Fluidigm ® Corporation (San Francisco, CA). This system is capable of running up to 2304 reactions per chip while only 96 total liquid loading steps are required. The advantages offered by this microfluidic chip, especially the large number of reactions that can be run in parallel, were beneficial to a study aimed at probing individual bacteria within the lignocellulose-decomposing microbial community that exists within wood-feeding termites [211] .

Within the literature, there exist several applications of microfluidic PCR amplification of specific pathogens. Some of these examples include; Dengue II virus [212] , E. coli SK [213] , Salmonella typhimurium [116, 214] , severe acute respiratory syndrome (SARS) [215, 216] , BK virus [217] , Campylobacter jejuni [218] , Human Papilloma virus (HPV) [219] , HIV [220] , Mycobacterium tuberculosis [221] , Influenza viral strain A/LA/1/87 [222] , Neisseria gonnorhoeae [223] , Bacillus anthracis [224] , Hepatitis B [225] , and Hepatitis C [226] . Another example, E. coli O157:H7, is a food-born pathogen that has been identified as the culprit in several severe human infection outbreaks [227] . This pathogen is particularly dangerous given that a minimal concentration is required for human infection (as low as 10-100 CFU/g) [228] , and that the organism produces many clinical manifestations [229] . It has been estimated that there are 73000 cases and 61 deaths per year within the United States alone that are attributable to this pathogen [230] . Traditional detection protocols, namely identifying the bacteria in stool cultures, can be time-intensive [230, 231] , requiring at least a 16 hour culture period [232] . In view of this limitation, Koh et al. have described a microfluidic device created from poly(cyclic olefin) for the amplification of E. coli O157:H7 and subsequent on-chip detection through electrophoresis and fluorescence identification of amplicons [116] . The static amplification was carried out within a gel-valve isolated reaction chamber with volumes as small as 29 nL using E. coli O157:H7 serotype specific primers. Heating was accomplished using a localized printed ink-based heater that allowed for a heating rate of ~12ºC/s. A detection limit of approximately 6 DNA copies present within the initial solution fed to the PCR was determined. Several additional examples of on-chip PCR of E. coli O157:H7 can be found within the literature [233, 234] .

There are several issues that complicate the rapid, microfluidic genomic detection of pathogenic substances. These concerns include low concentrations of biological targets and a typically complex biological sample matrix. Taken in context, there is thus some balance between minimum sample volume, rapid assay time, and detection sensitivity that must be considered in practical device fabrication. Considering this, Yang et al. amplified a DNA segment on-chip directly from an E. coli K12 strain (often referred to as colony PCR) [232] . The polycarbonate device employed a serpentine PCR channel for continuous amplification of volumes as large as 40 μl, and target DNA was provided through thermal lysis of the E. coli cells. The amplification was carried out using these cells both within and without the company of 2% sheep blood, and amplification from as few as 10 cells was observed. To block potential PCR inhibitors present within the biological sample matrix, the PCR mixture was supplemented with a specific buffer (Q buffer). Another potential means around this is to isolate the target genomic DNA after on-chip lysis. This was accomplished after thermal lysis of E. coli cells through capture onto probe tagged magnetic particles by Yeung et al. [235] . After DNA isolation, asymmetric PCR was performed to produce amplified single strand DNA that was subsequently detected through electrochemical methods. Other examples of microfluidic amplification of target DNA directly from a biological sample can be found within the literature [211, 217, 218, [236] [237] [238] .

There remain many key considerations regarding the miniaturization of PCR that must be resolved. Some of these platform design criteria can be seen in comparing the devices described by Koh et al. [116] and Yeung et al. [235] . This includes the PCR format, which can be static with the PCR mixture held within a reaction chamber while the chamber temperature is cycled through the normal temperature ranges, or continuous with the PCR mixture transported through a channel and different regions of the channel held to appropriate temperatures. Each of these designs present different challenges, such as achieving a rapid and precise temperature control and fluid manipulation. Additional considerations include providing for onchip heating and cooling (typically through either contact or non-contact methods), preventing sample evaporation (which is an increasing problem with a reduction in sample volume), ensuring sufficient sample volume to provide enough starting target DNA, preventing cross-contamination, preventing sample adsorption to surface walls, integrating sample preparation and detection platforms with the PCR on-chip, the material(s) employed for device fabrication, and providing for real-time PCR. As would be expected, the various materials available for construction of microfluidic devices each have different properties that will impact some aspect of the PCR reaction. Thus far, most microfluidic PCR devices have been fabricated utilizing either silicon or glass [116, 202, 239] . Silicon is advantageous for its high thermal conductivity and glass is advantageous because it lends itself well to integration with optical detection schemes. However, the use of these two materials presents specific challenges. For instance, bare silicon is opaque, complicating the integration of optical detection methodologies, and it is known to inhibit the PCR reaction through sample adsorption to the surface wall [239] . As is also the case with silicon chips, glass carries with it a high fabrication cost, preventing the realization of a truly disposable chip. This lack of disposability increases the risk of crosscontamination through multiple uses [226] . A potential means of accounting for cross-contamination involves control of the micro-chamber surfaces. To accomplish this end, Prakash et al. silanized a glass microchamber surface prior to carrying out the PCR reaction [220] . Following the reaction, the silanized surface of the microchamber was removed and a fresh silanization was carried out. In this manner, cross-contamination was controlled by ensuring sample antigen from a previous amplification did not remain on the wall of the microchamber during future amplification reactions. Another attractive solution to the cross-contamination problem involves the use of certain polymer substrates in the construction of microfluidic chips. Polymers, such as polydimethylsiloxane (PDMS), are relatively inexpensive, potentially allowing for the realization of a truly disposable chip, which would obviously circumvent the problem of crosscontamination. PDMS is also an attractive material as it provides good optical properties and exhibits limited PCR sample adsorption. Consequently there are several examples of the use of PDMS within microfluidic PCR designs [202, 213, 217, 239] . However, the use of PDMS is not without its own difficulties. One such difficulty is that PDMS is permeable, providing the possibility of sample losses. This problem becomes more significant as sample volume is reduced within a microfluidic chip. To avert this obstacle, the use of a vapor barrier has been suggested [240] . In one report, a polyethylene vapor barrier layer was implanted within the PDMS from which the device was fabricated [240] . Amplification was then carried out on 1.75 μl of PCR solution, with yeast genomic DNA as the target. The use of the vapor barrier was shown to reduce fluid loss by at least 3-fold, significantly enhancing the amplification. Finally it should not go without mentioning that sample evaporation is not a problem solely confined to polymer-fabricated devices, as this issue must also be considered when re-ducing sample volume within silicone and glass devices. Steps taken to minimize this sample loss include the use of mineral oil [209] and microvalves [241] . Recently, valve-less strategies based on diffusion-limited evaporation mechanisms for reducing sample evaporation were shown to be effective by Wang et al. [242] . To accomplish this aim, long and narrow channels were built into the microchip. As convective airflow is not present within these channels, sample evaporation then becomes limited by the lengthy diffusion time of vapor traversing the long, narrow channel. To supplement this strategy, efforts to decrease the liquid evaporation driving force included thermal isolation and vapor replenishment utilizing water.

The above discussion highlights only a few of the design issues facing the development of microfluidic PCR devices, and the reader is thus referred to several informative reviews that discuss this topic and also provide descriptions of the current state of microfluidic PCR technology [183, 201, 202, 239, [243] [244] [245] . Ultimately, it is unlikely that one particular design and design material will suffice for all potential uses of microfluidic PCR devices. It is expected that the final decision of which materials and designs to employ will be dependent on the specific intended use of the chip and the associated integrated components.

The goal of the research described in this chapter is development of more effective microfluidic diagnostic devices. The research generally proceeds along two inter-related fronts that can be viewed as being analogous to the concepts of hardware and software. The "hardware" aspect of this research is directed at designing better micro-devices (like pumps, sensors, etc.), formulating methods to fabricate these micro-devices in or on materials that are biocompatible (or desirable from some other standpoint), and in devising ways to package these systems and interface them with the outside (macro) world. The "software" aspect of this research is directed towards arriving at newer and more effective strategies for isolating and/or detecting target pathogens.

Thus far, most groups have been focused on different areas of "hardware" and "software" research, such as either developing improved biosensors for pathogen detection or developing new processes for the isolation and/or selective amplification of targets. While these efforts have produced very interesting and encouraging results, the development of a practical and ro-bust device that fulfils the pathogen detection goals laid out in the introduction of this chapter will likely require considerable focus on combining these varied approaches into an integrated microfluidic system. Such a system will conceivably (at least when describing first generation devices) employ a short pre-culture step in order to increase the concentration of detection targets. This step will be followed by relatively simple separation steps (perhaps utilizing phenomena such as DEP), and then the device will end with a sensitive biosensor of some type. In designing this system, it will be important to consider that the optimization of target pathogen detection (i.e. detecting low number of target pathogens with high specificity) might require the operation of the initial device stages (selective amplification and separation) sub-optimally.

The development of effective microfluidic devices for pathogen detection will also have to respond to new challenges that will arise from an increased understanding of the role microbes play within disease. For instance, recent work [246] has uncovered links between periodontal disease and the infestation of gums by archaea (a class of microscopic organism that is believed to have evolved separately from bacteria and other cellular organisms). Periodontal disease, in turn, has been identified as a contributing factor to endocarditis, atherosclerosis, stroke, and preterm delivery of low-birth-weight infants. Unfortunately, little is currently known about the exact mechanisms by which archaea could lead to these complications. Additionally, the protocols for quantifying the presence of such microscale organisms have either not been standardized, or are known only to a few expert researchers. Consequently, diagnostic devices (both micro-and macro-scale) that can successfully detect such microbes have yet to be developed.

Furthermore, diagnostic devices have largely been developed according to the conventional medical paradigm of "one disease, one pathogen." This model states that if a patient's symptoms match those of a particular disease, then the symptoms are expected to be caused by one type of pathogen (or a few related types) that have been previously associated with the respective disease in question. However, recent research has suggested that microbial "communities" may in fact either be responsible for and/or contributing towards several aggravating conditions, such as obesity [247] and Inflammatory Bowel Disease (IBD) [248, 249] . In such instances, the relative number, pathogen-host and pathogen-pathogen metabolic interaction, and spatial distribution of the micro-organisms likely determines their resulting impact (regarding both what the impact is and its severity) on the host. All these factors will add up to alter the technical goals required of microfluidic diagnostic devices and place increased demands on their capabilities in the future.

Emerging and re-emerging infectious diseases

The challenge of emerging and reemerging infectious diseases

Updating the accounts: global mortality of the 1918-1920 "Spanish" influenza pandemic

Guidelines for surveillance, prevention, and control of West Nile virus infection -United States

Infectious diseases: Considerations for the 21st century

Investigation of bioterrorism-related anthrax

Predictors of inappropriate antibiotic prescribing among primary care physicians

The relationship between the volume of antimicrobial consumption in human communities and the frequency of resistance

Threats to global health and survival: The growing crises of tropical infectious diseases -Our "unfinished agenda

Microfluidic diagnostic technologies for global public health

Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care

Heart Disease and Stroke Statistics--2006 Update. A Report From the American Heart Association Statistics Committee and Stroke Statistics Subcommittee

United States Cancer Statistics: 1999-2002 Incidence and Mortality Web-based Report Version

Blood culture in India: A proposal for a national programme for early detection of sepsis

Clinical impact of rapid invitro susceptibility testing and bacterial identification

Six-Year Prospective Study of Risk and Prognostic Factors in Patients with Nosocomial Sepsis Caused by Acinetobacter baumannii

Pseudomonas aeruginosa Bacteremia: Risk Factors for Mortality and Influence of Delayed Receipt of Effective Antimicrobial Therapy on Clinical Outcome

The Microbial world

Diversity of the Human Intestinal Microbial Flora

Investigation of Plaque Formation in Chick Embryo Cells as a Biological Marker for Distinguishing Herpes Virus Type 2 from Type 1

Quantitative Aspects of Septicemia

Production and characterization of a monoclonal antibody specific for enterohemorrhagic Escherichia coli of serotypes O157:H7 and O26:H11

Immobilization of immunoglobulins on silica surfaces

Fabrication of a capillary immunosensor in polymethyl methacrylate

Immobilization of DNA onto a polymer support and its potentiality as immunoadsorbent

Antibody Immobilization onto Glow Discharge Treated Polymers

Thiolated Salmonella antibody immobilization onto the gold surface of piezoelectric quartz crystal

MK (1995) Legionella contamination of dental-unit waters

Membrane-aerated microbioreactor for high-throughput bioprocessing

A microfluidic system for automatic cell culture

A Portable Anaerobic Microbioreactor Reveals Optimum Growth Conditions for the

Microchemostatmicrobial continuous culture in a polymer based, instrumented microbioreactor

Gas sorption, diffusion and permeation in poly(dimethylsiloxane)

2nd Annual International IEEE-EMBS Special Topic Conference on Microtechnologies in Medicine and Biology

Poly(dimethylsiloxane) (PDMS) and Silicon Hybrid Biochip for Bacterial Culture

Attachment and growth of bacteria on surfaces of continuous culture vessels

Long-Term Monitoring of Bacteria Undergoing Programmed Population Control in a Microchemostat

Encyclopedia of Microfluidics and Nanofluidics

Micromachined flow-through filter-chamber for chemical reactions on beads

Disposable integrated microfluidic biochip for blood typing by plastic microinjection moulding

Three-dimensional integrated microfluidic architectures enabled through electrically switchable nanocapillary array membranes

Polyimide microfluidic devices with integrated nanoporous filtration areas manufactured by micromachining and ion track technology

In-situ fabricated porous filters for microsystems

A new electro-osmotic pump based on silica monoliths

An electro-osmotic micropump based on monolithic silica for micro-flow analyses and electro-sprays

AIChE) Annual Meeting

Filter based microfluidic device as a platform for immunofluorescent assay of microbial cells

The 14th International Conference on Solid State Sensors, Actuators and Microsystems

AIChE) Annual Meeting

Magnetic design considerations for devices and particles used for biological high-gradient magnetic separation (HGMS) systems

Using magnetic beads to isolate inside-out glut1-containing vesicles from 3T3-L1 adipocyte plasma membranes

Continuous magnetophoretic separation of blood cells in microdevice format

A micropillar-integrated smart microfluidic device for specific capture and sorting of cells

One-step pathogen specific DNA extraction from whole blood on a centrifugal microfluidic device

Integrated reverse transcription polymerase chain reaction systems for virus detection

Combined microfluidicmicromagnetic separation of living cells in continuous flow

Magnetism and microfluidics

Magnetic separation in microfluidic systems using microfabricated electromagnets-experiments and simulations

Magnetic particle diverter in an integrated microfluidic format

Integrated microfluidic isolation platform for magnetic particle manipulation in biological systems

Applications of magnetic nanoparticles in biomedicine

Magnetic bead handling on-chip: new opportunities for analytical applications

Harnessing electrical forces for separation. Capillary zone electrophoresis, isoelectric focusing, field-flow fractionation, split-flow thincell continuous-separation and other techniques

Multistage Electrophoresis II: Treatment of a kinetic separation as a pseudo-equilibrium process

The Isoelectric Point of bacterial Cells

Electrokinetic Studies on Bacterial Surfaces: I. The Effects of Surface-active Agents on the Electrophoretic Mobilities of Bacteria

Hybrid microdevice electrophoresis of peptides, proteins, DNA, viruses, and bacteria in various separation media, using UVdetection

Capillary electrophoresis of biological particles: Viruses, bacteria, and eukaryotic cells

Handbook of Capillary Electrophoresis

Optimization of a rapid and sensitive identification system for Salmonella enteritidis by capillary electrophoresis with laser-induced fluorescence

Rapid CE Microbial Assays for Consumer Products That Contain Active Bacteria

Capillary Electrophoresis of Microbes

Separating Microbes in the Manner of Molecules: I. Capillary Electrokinetic Approaches

IEEE MEMS 2002 Las Vegas

Electrokinetic Mixing Vortices due to Electrolyte Depletion at Microchannel Junctions

Fluid Flow Induced by Non-Uniform AC Electrolytes on Micro-Electrodes: A Linear Double Layer Analysis

Pumping Liquids Using Asymmetric Electrode Arrays

Micro System Workshop '04

Dielectrophoretic characterization and separation of microorganisms

Insulator-based dielectrophoresis for the selective concentration and separation of live bacteria in water

Dielectrophoretic separation and manipulation of live and heat-treated cells of Listeria on microfabricated devices with interdigitated electrodes

2nd Annual International IEEE-EMB Special Topic Conference on Microtechnologies in Medicine & Biology

Particle separation by dielectrophoresis

Dielectrophoretic discrimination of bovine red blood cell starvation age by buffer selection and membrane cross-linking

Bovine red blood cell starvation age discrimination through an optimized dielectrophoretic approach with buffer selection and membrane cross-linking

Long-range AC Electrokinetic Trapping and Detection of Bioparticles

Aligning fast alternating current electroosmotic flow fields and characteristic frequencies with dielectrophoretic traps to achieve rapid bacteria detection

Rapid bioparticle concentration and detection by combining a discharge driven vortex with surface enhanced Raman scattering

Electrokinetic particle aggregation patterns in microvortices due to particle-field interaction

An integrated highthroughput dielectrophoretic chip for continuous bio-particle filtering, focusing, sorting, trapping, and detection

Dielectrophoresis switching with vertical sidewall electrodes for microfluidic flow cytometry

A 3-D microelectrode system for handling and caging single cells and particles

Surface micromachined dielectrophoretic gates for the front-end device of a biodetection system

AC Electrokinetics: colloids and nanoparticles

Combined fieldinduced dielectrophoresis and pahse separation for manipulating particles in microfluidics

Dielectrophoretic sorting of particles and cells in a microsystem

High throughput particle analysis: Combing dielectrophoretic particle focussing with confocal optical detection

Bacteria capture, concentration and detection by AC dielectrophoresis and self-assembly of dispersed single wall carbon nanotubes

Dielectrophoretic trapping of single bacteria ar carbon nanofiber nanoelectrode arrays

Impedimetric monitoring of cell attachment on interdigitated microelectrodes

Selective stection of bacteria by a dielectrophoretic impedance measurement method using an antibodyimmobilized electrode chip

Targeted cell detection based on microchannel gating

Electrical Detection of Single Viruses. Proceedings of the National Acadacemy of Science

Measurement of charge transfer during bacterial adhesion to an indium tin oxide surface in a parallel plate flow chamber

Medium and Interface Components in Impedance Microbiology

A micro-scale multi-frequency reactance measurement technique to detect bacterial growth at low bioparticle concentrations

Microfluidics-based optical biosensing method for rapid detection of Escherichia coli O157 : H7

A microfluidic filter biochip-based chemiluminescence biosensing method for detection of Escherichia coli O157 : H7

Convergence of quantum dot barcodes with microfluidics and signal processing for multiplexed high-throughput infectious disease diagnostics

Integrating polymerase chain reaction, valving, and electrophoresis in a plastic device for bacterial detection

Development of a nanoparticle-labeled microfluidic immunoassay for detection of pathogenic microorganisms

The first SERRS multiplexing from labelled oligonucleotides in a microfluidics lab-on-a-chip

Fast, ultrasensitive virus detection using a young interferometer sensor

Optofluidic integration for microanalysis

Unconventional detection methods for microfluidic devices

Optical and electrochemical detection techniques for cellbased microfluidic systems

Recent developments in optical detection methods for microchip separations

Pathogen detection: A perspective of traditional methods and biosensors

Microfluidic immunosensor systems

Optical sensing systems for microfluidic devices: A review

Miniaturized immunoassay microfluidic system with electrokinetic control

Membrane-based on-line optical analysis system for rapid detection of bacteria and spores

Monolithic capillary electrophoresis device with integrated fluorescence detector

World-to-chip microfluidic interface with built-in valves for multichamber chip-based PCR assays

Real time PCR on disposable PDMS chip with a miniaturized thermal cycler

Single-molecule fluorescence detection in microfluidic channels -the Holy Grail in mu TAS?

Bio-assay based on single molecule fluorescence detection in microfluidic channels

Microscale bioanalytical systems

A handheld real time thermal cycler for bacterial pathogen detection

color, 10 parameter flow cytometry to elucidate complex leucocyte heterogeneity

Current status of flow cytometry in cell and molecular biology

Microfluidics for flow cytometric analysis of cells and particles

Engineering flows in small devices: Microfluidics towards lab-on-a-chip

Electrokinetics: a viable microfluidic platform for miniature diagnostic kits

A review of micropumps

Rapid and simple quantification of bacterial cells using a microfluidic device

An integrated microfluidic system for reaction, high-sensitivity detection, and sorting of fluorescent cells and particles. Analytical Chemistry 75

Flow cytometry of Escherichia coli on microfluidic devices

An Integrated Microfabricated Cell-sorter

International Society for Applied Cytology XXII International Congress

Dynamic response of a cantilever in liquid near a solid wall

Detection of bacterial cells and antibodies using surface micromachined thin silicon cantilever resonators

Rapid Detection of Foodborne E. coli O157:H7 Using Piezoelectric-excited Millimeter-size Cantilever Sensors

Micromechanical cantilevel-based biosensors

Mechanical resonant immunospecific biological detector

Design, fabrication and testing of a novel MEMS resonator for mass sensing applications

System on chip mass sensor based on polysilicon cantilevers arrays for multiple detection

Rapid molecular detection of food and water-borne diseases

Resonators with integrated CMOS circuitry for mass sensing applications, fabricated by electron beam lithography

7th International Conference on Solid-State Sensors and Actuators (Transducers '93)

Surface adhesion reduction in silicon microstructures using femtosecond laser pulses

93rd annual meeting of the US Animal Health Association, Las Vegas

Rapid detection of total and fecal coliforms in water by enzymatic hydrolysis of 4-methylumbelliferone-beta-D-galactoside

Early detection of bacterial growth and carbon 14 labelled glucose

A Rapid Microtiter Plate Method To Measure Carbon Dioxide Evolved from Carbon Substrate Amendments so as To Determine the

Electronic measurement of bacterial growth

Comparative evaluation of different chromogenic/fluorogenic media for detecting Escherichia coli O157:H7 in food

Experience with ESP Liquid Culture System and the Rapid Culture Identification of Mycobacterium avium subsp. paratuberculosis (MAP)

Are growth rates of Escherichia coli in batch cultures limited by respiration

Enumeration and biomass estimation of planktonic bacteria and viruses by transmission electron microscopy

Impedance Microbiologyon-a-Chip: Microfluidic Bioprocessor for Rapid Detection of Bacterial Metabolism

Conductivity and pH Dual Detection of Growth Profile of Healthy and Stressed Listeria monocytogenes

Designing a nano-interface in a microfluidic chip to probe living cells: Challenges and perspectives

Calorimetry of microbial growth using a thermopile based microreactor

Integrated cell positioning and cell based ISFET biosensors

Advances in nucleic acid-based diagnostics of bacterial infections

Nucleic acid-based methods for the detection of bacterial pathogens: Present and future considerations for the clinical laboratory

Immunoassay, DNA analysis, and other ligand binding assay techniques: From electropherograms to multiplexed, ultrasensitive microarrays on a chip

Development and characterization of an ELISA assay in PDMS microfluidic channels

Pathogenesis and diagnosis of shiga toxinproducing Escherichia coli infections

Detection of preformed type A botulinal toxin in hash brown potatoes by using the mouse bioasssay and a modified ELISA test

Development of a complete ELISA using Salmonella lipopolysaccharides of various serogronps allowing to detect all infected pigs

Laboratory testing for infection with the human immunodeficiency virus: Established and novel approaches

Antibody arrays: an embryonic but rapidly growing technology

Recent progress in microfluidic devices for nucleic acid and antibody assays

Detection of E-coli using a microfluidics-based antibody biochip detection system

Surface modification for enhancing antibody binding on polymer-based microfluidic device for enzyme-linked immunosorbent assay

Development of a novel electrokinetically driven microfluidic immunoassay for the detection of Helicobacter pylori

Microfluidic device as a new platform for immunofluorescent detection of viruses

Integration of an immunosorbent assay system: Analysis of secretory human immunoglobulin A on polystyrene beads in a microchip

Bead-assisted displacement immunoassay for staphylococcal enterotoxin B on a microchip

Enzyme-linked Immunosorbent assay of Escherichia coli O157 : H7 in surface enhanced Poly(Methyl methacrylate) microchannels

Solid supports for micro analytical systems

Microfluidic immunoassay platform using antibody-immobilized glass beads and its application for detection of Escherichia coli O157 : H7

Multiplex PCR assay for rapid detection and genotyping of Helicobacter pylori directly from biopsy specimens

Comparison of the amplified Mycobacterium tuberculosis (MTB) direct test, Amplicor MTB PCR, and IS6110-PCR for detection of MTB in respiratory specimens

Novel technique of quantitative nested real-time PCR assay for Mycobacterium tuberculosis DNA

Direct detection of Mycobacterium tuberculosis in clinical specimens using single-tube biotinylated nested polymerase chain reaction-enzyme linked immunoassay (PCR-ELISA)

Multicenter evaluation of AMPLICOR and automated COBAS AMPLICOR CT/NG tests for Neisseria gonorrhoeae

Multicenter study of a rapid molecular-based assay for the diagnosis of group B Streptococcus colonization in pregnant women

Development of procedures for rapid detection of E-coli O157 : H7 from source and finished water samples

Integrated genetic analysis microsystems

PCR microfluidic devices for DNA amplification

Tranducer '93, seventh international conference on solid state Sens Actuators

Integrated portable genetic analysis microsystem for pathogen/infectious disease detection

Nanoliter high throughput quantitative PCR

Parallel picoliter RT-PCR assays using microfluidics

Microchamber array based DNA quantification and specific sequence detection from a single copy via PCR in nanoliter volumes

Planar chip device for PCR and hybridization with surface acoustic wave pump

Ultra fast miniaturized real-time PCR: 40 cycles in less than six minutes

Rapid and sensitive detection of Mycobacterium DNA using cepheid SmartCycler (R) and tube lysis system

Microfluidic digital PCR enables multigene analysis of individual environmental bacteria

Integrated polymerase chain reaction chips utilizing digital microfluidics

DNA amplification on a PDMS-glass hybrid microchip

On-chip pressure injection for integration of infrared-mediated DNA amplification with electrophoretic separation

Microfluidic handling of PCR solution and DNA amplification on a reaction chamber array biochip

Determination of SARScoronavirus by a microfluidic chip system

Automated screening using microfluidic chip-based PCR and product detection to assess risk of BK virus-associated nephropathy in renal transplant recipients

Detection of a putative virulence cadF gene of Campylobacter jejuni obtained from different sources using a microfabricated PCR chip

Droplet-based micro oscillating-flow PCR chip

An integrated genetic analysis microfluidic platform with valves and a PCR chip reusability method to avoid contamination

An integrated nanoliter DNA analysis device

An integrated microfluidic device for influenza and other genetic analyses

Chemical Amplification: Continuous-Flow PCR on a Chip

A continuous-flow polymerase chain reaction microchip with regional velocity control

Clinical evaluation of microscale chip-based PCR system for rapid detection of hepatitis B virus

A microchip-based PCR device using flexible printed circuit technology

Evaluation of culture methods to identify bovine feces with high concentrations of Escherichia coli O157

Survival of Escherichia coli O157 : H7 and Salmonella enterica serovars Typhimurium in iceberg lettuce and the antimicrobial effect of rice vinegar against E.coli O157 : H7

Sensitivity of two techniques to detect Escherichia coli O157 in naturally infected bovine fecal samples

Real-time detection of Escherichia coli O157 : H7 sequences using a circulating-flow system of quartz crystal microbalance

High sensitivity PCR assay in plastic micro reactors

Real time PCR on disposable PDMS chip with a miniaturized thermal cycler

Electrokinetically controlled realtime polymerase chain reaction in microchannel using Joule heating effect

A DNA biochip for on-thespot multiplexed pathogen identification

Infectious disease -PCR detection of bacteria in seven minutes

Microchip module for blood sample preparation and nucleic acid amplification reactions

Microchip device for cell lysis, multiplex PCR amplification, and electrophoretic sizing

Miniaturized PCR chips for nucleic acid amplification and analysis: latest advances and future trends

Small volume PCR in PDMS biochips with integrated fluid control and vapour barrier

Fully integrated PCR-capillary electrophoresis microsystem for DNA analysis

Microfabricated valveless devices for thermal bioreactions based on diffusion-limited evaporation

Advances in polymerase chain reaction on microfluidic chips

The micro-total analytical system for the detection of bacteria/viruses

Miniaturised nucleic acid analysis

Methanogenic Archaea and Human Periodontal Disease

Obesity alters gut microbial ecology

Current concept on the pathogenesis of inflammatory bowel disease -crosstalk between genetic and microbial factors: Pathogenic bacteria and altered bacterial sensing or changes in mucosal integrity take "toll

Role and regulation of intestinal epithelial heat shock proteins in health and disease