FOR DETECTING CONTAMINANTS

Cross-Reference to Related Application

This application claims priority to co-pending U.S. Provisional Application Serial

Number 62/675,515, filed May 25, 2018, which is hereby incorporated by reference herein in its entirety.

Background

Ever since the discovery of penicillin, resistance to antimicrobial agents in pathogenic bacteria has been steadily rising in a multitude of areas, from food to clinical cases. This rise is associated with intensive use of antibiotics in livestock and in human health care. The lack of effective antibiotic therapies against antibiotic- resistant bacteria leads to prolonged treatments, increased morbidity and mortality, together with inflated health care costs. Developing novel antibiotics that kill resistant bacteria is no longer a viable solution since pathogenic bacteria have found ways to evade all new antimicrobial compounds regardless of the drugs’ mechanisms of action. Therefore, strategies that effectively use existing and future antibiotics, while integrating an in-depth knowledge of the evolutionary dynamics of antibiotic resistance, are urgently needed.

Effective treatment of infectious diseases caused by pathogenic bacteria often requires early detection of bacteria and administering the most effective antibiotics based on the available information about the disease-causing pathogens. Despite all the technological advances in medicine, detection and characterization of pathogenic bacteria can still take several days. In many cases, physicians have to start antibiotic treatments based on anecdotal experience, leading to biases in pathogen diagnosis and isolation, and dramatic variation in the way physicians treat these infections. Depending on the prognosis of the infectious disease, patients are given several antibiotics, which can lead to dysbiosis in healthy microbiota, further exacerbating the antibiotic resistance problem in hospital settings. Therefore, there is a need for a rapid, affordable, and highly sensitive system and method that can reduce the time needed for detection and measurement of antibiotic sensitivity below several hours and have point of care (ROC) deployment capability.

Brief Description of the Drawings

The present disclosure may be better understood with reference to the following figures. Matching reference numerals designate corresponding parts throughout the figures, which are not necessarily drawn to scale.

Fig. 1 is a schematic view of an embodiment of a contaminant detection system that utilizes the high reflectance coefficient of light at the interface between thin (low refractive index) and dense (high refractive index) media at high incidence angles.

Fig. 2 is a detail view of the hollow core of a hollow tube of the detection system of Fig. 1 , showing the focal distance of the laser being set such that the incidence angle (Q) of the laser at the tube wall remains greater than 89 degrees. At Q > 89°, the reflectance coefficient of laser light is R > 0.82 for both 5 and p polarizations of light.

Fig. 3 is a schematic view illustrating light projected into the hollow tube of the detection system. Photons of the light encountering a cell or a particle are either absorbed or scattered losing the angular condition to be guided by the hollow tube

(i.e., Q > 89°). Any cell on the light path reduces the total intensity proportional to its cross-sectional area and creates a two-dimensional shadow-projection image on the optical detector.

Fig. 4 is a graph that shows a calibration between the voltage signal of the light detector and cell-density data. Simulation data based on a disclosed model (line) and calibration data (open circles) for the detection system agrees well (R ² = 0.97) up to an optical density (OD) < 10 ^-2 and ( R ² = 0.93) for the whole data range used.

Fig. 5(a) is a graph that shows a simulated comparison of a signal based on the detection system model compared with Beer-Lambert at the low OD regime for hollow tube dimensions / = 400 mm and d = 500 pm for a bacterial cell diameter of 1 pm.

Fig. 5(b) is a graph that shows signal strength versus number of cells and cell dimensions in the detection system. As the cell size increases, the signal strength is multiplied proportionally to the cell cross-sectional area. The cell shape is taken as a circular pattern with the specified dimensions. The region marked with asterisks indicates the boundary of detection where the signal is above the 1 mV limit. The signal rapidly saturates over OD ~ 0.1 as cells optically block the hollow tube cross-section.

Fig. 5(c) is a graph that shows the dependence of signal strength on hollow tube inner diameter at a constant number of cells for Escherichia coli (E coii).

Increased cell size and reduced hollow tube diameter employs the same principle that contributes to enhanced sensitivity. Fig. 5(d) is a graph that shows that a t-test analysis for sensitivity is significant with a level of p = 0.02 compared to blank media even for OD levels of 10 ^"7 Calibrated

OD levels are confirmed via colony counts.

Fig. 6(a) is a graph that shows continuous growth curves for the same bacterial culture in the disclosed detection system and optical plate reader systems started from OD ~ 5x10 ^"7 . In data acquired by the plate reader, the lag phase lasts about 7 hours whereas the disclosed detection system indicates immediate growth in contrast to the accepted bacterial growth phase model.

Fig. 6(b) is a graph that shows a growth curve for two different cultures taken using the disclosed detection system. Notice the continuous and immediate growth phase with the initiation of each culture. The growth rate for culture A1 was 1.707 hour

1 and the growth rate for culture A2 was 1.715 hour ¹ . The difference between the growth curves is related to the initial number of cells in the hollow tube volume. At OD ~ 10 ^'7 , the expected number of bacteria within the hollow tube volume is about 4. Even a single bacterium can introduce a delay that is equal to the generation time between two cultures. The doubling time for both cultures is approximately 24 minutes.

Figs. 7(a) and 7(b) are graphs that show the results of E. coli antibiotic susceptibility testing with (a) the disclosed detection method versus (b) conventional methods. Parallel growth measurements of the same bacterial culture in the disclosed detection system and a spectrometer demonstrate that the disclosed detection system can provide growth data as early as 2 hours. Dose-response measurements with the system can be obtained by the second hour of monitoring in comparison to conventional methods for Piperacillin-Tazobactam in which a spectrometer requires 7 hours to provide the first dose-response data. The asterisks indicate if a minimum inhibitory concentration (MIC) determination is possible based on the data. The insets show dose-response curves for highlighted regions when MIC was first identifiable in each case.

Figs. 8(a)-8(c) are graphs that show cell counts detected using the inventive system for each of C. glabrata, S. cenevisiae, and E. Coli, respectively.

Figs. 8(d)-8(g) are graphs that show optical densities of bacterial cultures grown in lysogeny broth (LB) media (started from OD ~ 5 x 1 O ^'7 ) and recorded by the inventive system for each of E. coli, S. aureus, P. aeruginosa, and an E. coli clinical isolate, respectively.

Detailed Description

As described above, there is a need for a rapid, affordable, and highly sensitive system and method that can reduce the time needed for detection and antibiotic sensitivity below several hours and have point of care (ROC) deployment capability.

Disclosed herein are embodiments of such a system and method. In one embodiment, a system comprises a small hollow tube into which a liquid sample to be analyzed can be delivered. Laser light is directed into the hollow core at one end of the hollow tube at an extremely shallow angle that is nearly parallel with the longitudinal axis of the hollow tube. When contaminants, such as bacterial cells, are present in the sample, photons encountering a cell are either scattered, losing the angular condition to be guided by the hollow tube (i.e., Q > 89°), or absorbed by the contaminants, reducing the intensity of the light that reaches a light detector positioned at the opposite end of the hollow tube. This reduction in light intensity is indicative of the presence and concentration of the contaminants within the sample. In some embodiments, the system can detect extremely low cell densities (optical density (OD) >10 ^"7 ) in a manner of seconds. This corresponds to less than 50 bacterial cells in 1 ml of liquid for E. coli. The sensitivity of the system is multiple orders of magnitude higher than standard optical density reading instruments that rely on light absorption or scattering and the system can be produced for very little cost.

In the following disclosure, various specific embodiments are described. It is to be understood that those embodiments are example implementations of the disclosed inventions and that alternative embodiments are possible. All such embodiments are intended to fall within the scope of this disclosure.

Fig. 1 illustrates an embodiment of a system 10 for detecting contaminants, such as bacterial cells, in a liquid sample. As shown in this figure, the system 10 generally comprises a hollow tube 12 having a hollow core that is utilized as an optofluidic channel. In some embodiments, the hollow tube 12 is a fused-silica optical fiber having an inner core diameter of approximately 100 pm to 500 pm (e.g., 500 pm), a length of approximately 50 mm to 400 mm (e.g., 400 mm), and walls having an index of refraction, n, of approximately 1.40 to 1.67 (e.g., 1.458). An example of such an optical fiber is part number 25739 produced by Supelco.

Each end of the hollow tube is 12 is capped with an end cap 14 that, for example, can be made of a polymeric material, such as polydimethylsiloxane (ROMS).

As shown in the inset detail drawing, each end cap 14 includes a first passage 16 that is configured to receive an end of the hollow tube 12 and a second passage 18 that is configured to receive an end of an inlet 20 or an outlet 22, depending upon the end of the hollow tube with which the end cap is associated. Also visible in the detail drawing, is an inner cavity 24 of the end cap 14 that is in fluid communication with each of the passages 16, 18. As can be appreciated from the drawings, the inlet 20 and outlet 22 are orthogonal to the hollow tube 12 and, therefore, extend away from the hollow tube at a 90° angle. As can also be appreciated from the drawings, fluid can flow through the inlet 20, into and through the hollow tube 12, and then into and through the outlet

22. Attached to each end cap 14 is an end plate 26 that, for example, can be made of a glass material. The end plates 26 form a direct optical path into the hollow core of the hollow tube 12. In some embodiments, the end plates 26 can be fixedly attached to the end caps 14.

With further reference to Fig. 1 , a laser light source 28 can be used to deliver laser light into a first end of the hollow core of the hollow tube 12. In the example of

Fig. 1 , this laser light is introduced into the hollow core at the inlet end of the hollow tube 12. By way of example, the laser light source 28 comprises a laser diode assembly including a 5 mW, 650 nm laser diode and a focusing lens, such as a lens having part number 0710893-000 produced by AixiZ. Significantly, the laser light is emitted at a shallow or“grazing” angle that is nearly parallel with the longitudinal axis of the hollow core of the hollow tube. As shown in Fig. 2, which comprises a schematic detail view of the hollow core of the hollow tube 12, this angle, Q, can be greater than 0° but less than 1 °. Stated otherwise, the angle of incidence of the light relative to the inner walls of the hollow tube is between 89° and 90°.

Positioned at the opposite end (i.e., the outlet end in Fig. 1 ) of the hollow tube

12 is a light detector 30. The light detector 30 can, for example, comprise a photodiode having a peak sensitivity at 940 nm (e.g., Digi-key 1080-1 148-ND). The light detector 30 outputs a current that is converted to a voltage, which is indicative of the light intensity detected by the light detector. The voltage readings can be recorded by an appropriate data acquisition device (not shown), such as a data acquisition card (MCC,

1408 HS), and the readings can be processed using a software program comprising one or more algorithms including computer-executable instructions, such as Matlab, that may be executed on a computing device (not shown), such as a desktop computer, that comprises a processor and a computer-readable storage medium.

As noted above, the hollow core is utilized as an optofluidic channel, which serves as a selective waveguide for the laser light and as a detection/growth chamber for contaminants, such as bacterial cells. Such cells can be delivered to the optofluidic channel either via continuous circulation, which can include a degassing/bubble trap mechanism to avoid bubble formations, or by manual injection for cell density measurements.

The theoretical basis for the above-described detection system will now be discussed. Generally speaking, the system takes advantage of both optical absorption and scattering events in the detection process. The laser beam is coupled into the hollow core of the hollow tube, together with cells in a growth medium (e.g., index of refraction n = 1.33). The focal distance of the laser is set such that the incidence angle of the laser at the hollow tube wall remains at a grazing angle that is less than 90° but greater than 89°. Significantly, this maintains the reflectance coefficient R > 0.82 for both s and p polarizations of light, as depicted in Fig. 2. Light that hits obstacles on the travel path loses the condition for high reflectance through absorption or scattering and is selectively filtered out along the hollow tube, as depicted in Fig. 3. This creates a two-dimensional cross-sectional projection image of the cells dispersed inside a three-dimensional volume at the end of the hollow tube. In this projection image, the total light lost during travel is given by: where Io is the intensity of light, a, is the cross-sectional area of the z ^' th bacterial cell, n is the number of bacterial cells, and r is the average diameter of a cell (assuming a spherical shape). The image is transformed to a voltage signal through the light detector, where the difference in reference to initial blank media measurement is converted to optical density (OD = 1 corresponds to 5 x 10 ⁸ cells per ml of culture) readings using the calibration curve in Fig. 4.

In conventional absorption measurements that employ the Beer-Lambert relation, absorption is given as: where m is the attenuation coefficient and l is the length of the specimen. For a well dispersed absorbing species, m is also dependent on the volumetric concentration. For a non-uniform absorption case, as in bacterial cultures, light absorption per cell is correlated with the size of the cell and is undetectable. Therefore, optical density measurements for bacterial cells do not typically yield statistically significant signals.

A physical model explaining the working principle of the disclosed system suggests that the AF response is proportional to the ratio of projected bacterial area and hollow tube cross-sectional area. Thus, increased average bacterial size or decreased hollow tube diameter enhances sensitivity. At a minimum sensitivity threshold of AV = 1 mV, the detection system is capable of sensing about 27 bacteria of 1 pm, or a single cell of 8 pm diameter in a standard 500 pm diameter, 400 mm long hollow tube setup (Figure 5(b)).

Signal strength (AF) behaves in accordance with AVa l IIP where R is the inner diameter of the hollow core. Thus, a decrease in diameter enhances the sensitivity per bacterial cell, as in Figure 5(c), where DG increases from 0.002 mV to 80 mV at a constant number of bacteria for a 400 mm long hollow tube. On the other hand, the projection coverage of the bacteria increases linearly with the encapsulated hollow tube volume laying out a simple AVaL relation with the length of the hollow tube.

The signal in the detection system model saturates with the bacterial density at relatively low cell density levels (OD ~ 0.01 ) because the narrow cross-section of a hollow tube is optically blocked by a rather low number of randomly distributed bacteria, even at low densities. Fig. 5(b) shows the diminishing dynamic response of the detection system by increasing the number of cells.

The minimum sensitivity level attained with the 400 mm long, 500 pm diameter setup is OD >10 ⁷ for E. co/z (MG1655), as shown in Fig. 5(d). The measurements start from blank media as the reference point, where the culture used at each data point is serially diluted from an E. co/z culture (OD = 10 ² ) measured with a spectrometer. The measurements for OD ~ 10 ^-7 density yields a significance level of p = 0.02 in comparison to blank media, and starting from OD > 10 ^-® density, the measurements are highly significant with levels of p « 0.001. The diluted cultures used for the measurements were also confirmed by colony forming unit (CFU) counts conducted for the same bacterial cultures.

One of the most promising applications for the disclosed detection system is in the measurement of bacterial growth rate and monitoring the course of the growth at very low bacterial concentrations. Depending on the growth conditions and cell type, conventional techniques typically require time scales between seven hours to several days to characterize the growth. The disclosed detection method provides enough data for rapid detection of growth in one hour, which may be extremely important for critical circumstances in clinical settings. Figs. 6(a) and 6(b) demonstrate the capability of the detection system in growth detection. Bacterial growth is typically modeled after four different phases: lag phase, exponential phase, stationary phase, and death phase. Lag phase is considered the adaptation period required for bacterial cells to facilitate to new environmental conditions. There is no scientific consensus on the definition and duration of lag phase, and the start and finish of the phase is vague because conventional methods fail to monitor cell growth at very low densities.

The advantage of the disclosed detection system in measuring the growth rate of bacterial cultures at extremely low densities tells the story from a different perspective. At these densities, growth rate measurements performed with standard plate readers at room temperature (T = 23°C) for the same bacterial culture with a starting cell density of OD = 5x1 O ^'7 shows no growth up to the 7th hour, whereas parallel detection system measurements indicate bacterial growth within the first hour, as shown in Fig. 6(a). Therefore, the lag phase as it is defined can be a pseudo-phase definition for some cases that arise due to the lack of monitoring capability at low cell densities. Indeed, from measurements with the disclosed detection system shown in

Fig. 6(b), bacteria show immediate and continuous growth, even at the very initial stages in two separate measurements taken at intervals of approximately 20 minutes at T = 37°C. The doubling time for these two measurements is very similar at 24 minutes, but the course of growth may vary because of the inherent phenotypic stochasticity at very low cell densities.

Antibiotic susceptibility testing (AST) is crucial in diagnostic laboratories to test the sensitivity of a particular bacterial strain to an antimicrobial agent and to monitor general trend of antibiotic resistance evolution. Existing antibiotic susceptibility tests are based on three methods: disk diffusion, gradient diffusion, and agar/broth dilution.

The time-consuming factor to these methods is the overnight incubation of the bacterial cultures necessary for detectable growth. In comparison to many other approaches, the disclosed method provides critical information about the bacteria affecting a patient in a few hours and could help ease the rising challenge of drug- resistant bacteria strains. In order to illustrate the potential advantages of the disclosed detection method, a susceptibility test was performed for the antibiotic piperacillin- tazobactam (P-T) for E coli (MG1655) in the disclosed detection system and a spectrometer in parallel (initial cell density = ~150 cells per ml or ~3 x 10 ^-7 ).

As shown in Fig. 7, the disclosed detection system is capable of revealing the complete character of the growth-dose dependence within the first approximately 2 hours of growth. The minimum inhibitory concentration (MIC) value (> 2 pg/ml) can be determined within this period without further need to continue the test. In contrast, a conventional measurement in a microtiter plate requires over 7 hours to yield a comparably reliable MIC value. Thus, for the case of E. coli and piperacillin- tazobactam, the disclosed detection system reduces the time for susceptibility characterization by 5 hours, which can be critical for clinical cases. This can cut the time required from disease monitoring to intervention by reducing the diagnosis and decision-making cycles further in other species of bacteria in life-threatening conditions.

Further experiments were performed to evaluate the sensitivity of the disclosed system. The system was first calibrated using bacterial (E coli, Staphylococcus aureus, and Pseudomonas aeruginosa) and fungal ( Saccharomyces cerevisiae,

Candida glabrata) cells. Overnight cultures were serially diluted and the voltage signals generated (DNZ) in in the system were recorded. The cell densities of diluted cultures were also confirmed by colony forming unit (CFU) counts conducted for the very same bacterial cultures (1 OD ~ 5 x 10 ⁸ CFU/mL for all tested bacteria) and fungal cultures (1 OD ~ 2.2 x 10 ⁷ CFU/mL for S. cerevisiae and 1 OD ~ 1.25 x 10 ⁷ CFU/mL for C. g/abrata). The measured DNZ values and corresponding OD values were linearly correlated on a logarithmic scale on both axes. Based on this linear correlation, a power function (Iog ₁₀ (AV) = a + b log ₁₀ (OD)) was used to fit the data and convert voltage readings to OD values and cell counts. In the experimental settings for bacteria and fungi, the value of a was typically between 2.9 and 3.1 whereas b, the slope of the calibration curve, was found to be between 0.9 and 1.1. The limit of detection (LOD) attained at the standard 400 mm long, 500 pm setup was OD 5 x 10 ⁷ for bacteria which corresponds to approximately 250 bacterial CFU/mL. The limit of quantification (LOQ, minimum level of signal after which detector response increases proportional with the cell density) for the system was OD ³ 1.5 x 10 ^-6 for bacteria, which corresponds to approximately 750 bacterial CFU/mL.

Given that the detection volume was approximately 80 pL, as low as approximately 20 bacterial cells can be detected, which were measured to be approximately 10 ⁴ -fold more sensitive than a commercial plate reader (TECAN M200

Pro). Remarkably, it was determined that the system can detect a single fungal cell

(spherical S. cerevisiae or C. glabrata with a 5-6 pm diameter). All measurements with bacterial cells were performed in lysogeny broth (LB) media. The system is, as all optical systems are, limited by the transparency of the medium to the laser light that is used. Using a more transparent media, such as minimal M9 media, improves the sensitivity of the system by increasing signal to noise ratio, especially in longer hollow tube schemes decreasing the LOD down to ~2 x 10 ⁷ , which corresponds to approximately 8 to 10 bacterial cells. For practical MIC measurement purposes, any transparent media commonly available can be utilized in measurements. Fig. 8 presents various cell-count results. The top panel of Fig. 8(a) shows cell counts for C. glabrata at OD approximately 10* and displays a Poisson distribution with counts of 0, 1 , 2, and 3 with a mean value of approximately 1. The bottom panel shows that increasing the yeast cell density shifts the distribution towards a normal distribution with a mean value of approximately 5.3 cells. The top panel of Fig. 8(b) shows cell counts for S. cerevisiae at an OD nearly 10* which results in a Poisson distribution with counts of 0, 1 , and 2 and a mean value of approximately 0.48. The lower panel shows that a slight increase of cell density moves the distribution towards a normal distribution with a mean value of approximately 4 cells. The top panels of Fig. 8(c) shows that the cell counts for an E. coli culture at an optical density of approximately 5x110 ^-7 yields approximately 25 cells. The lower panel shows that increasing the bacterial cell density slightly shifts distribution towards a normal distribution with a similar mean value of approximately 24 cells with smaller standard deviation, which shows the sensitivity limit of the system. The optical densities of bacterial cultures grown in LB (started from OD ~ 5 x 10 ^'7 ) and recorded by the system are shown for each of £. ¥li (MG1655) (Fig. 8(d)), S. aureus (RN4220) (Fig. 8(e)), P. aeruginosa (PA01 ) (Fig. 8(f)), and an £. coli clinical isolate (ET-CI28) (Fig. 8(g)).

An antibiotic susceptibility assay was devised to compare the inventive method with the clinically approved standard antibiotic susceptibility tests, which typically start from higher cell densities (i.e., OD > 10~ ³ ). A commercially available Gram negative

MIC panel (NM43, Beckmann Coulter B1017-420) was utilized, which includes 29 antibiotics and antibiotic combinations (Table 1 ). Ten clinical £. coli isolates (CIET-

001 -010) that were previously phenotyped at the Clinical Microbiology Laboratory of

Children’s Medical Center (CMLCMC), Dallas, TX, were used. The manufacturer's protocol for preparation and inoculation of the MicroScan Neg MIC Panel Type 43 (NM43, Beckman Coulter) was followed in which a starting cell density of 0.5

McFarland (about 0.25 ODeoo) is then further diluted before inoculating into the antibiotic panel with a final target inoculum of 3 to 7x10 ⁵ CFU/ml. First, every clinical strain was grown for 24 hours and MIC values were determined by visual inspection following the guidelines provided by the manufacturer. It was confirmed that the MIC values that were obtained were in line with the MIC values provided by the CMLCMC.

Then, one of the clinical strains (CIET-001 ) that was resistant to some of the compounds in the antibiotic panel was grown using different incubation times for optimizing the minimum inhibition time required for MIC determination with the inventive system.

Ten consecutive cell density measurements were performed using the system at hourly intervals and the growth in each well of the NM43 MIC test panel was tracked.

Using three different growth thresholds (half, one, and two doublings compared to initial inoculum size of OD ~ 2 x 10 ^"3 ), growth curves were converted to binary resistance maps. This data set was compared to the expected resistance map for the same E. coli isolate. Correlations between the observed resistance map at different time points and the expected resistance map were calculated using the Mathews

Correlation Coefficient true positive, FP:

false positive, TN: true negative, and FN: false negative). How this correlation changed as a function of growth threshold was also evaluated. It was determined that for this particular strain (CIET-001 ), when a growth threshold of half doubling was considered, waiting three hours or longer was enough to precisely reproduce the MIC values we obtained by 24 hour incubation. Therefore, for consistency and improved accuracy, the growth threshold was set to one doubling and incubation time to 4 hours. Clinical isolates (CIET-002-CIET-010) were tested by taking single measurements after 4 hours of incubation. The correlation coefficient is 1.00 (perfect correlation) for

7 strains tested, with the remaining 3 isolates exceeding 0.9. A growth map for all 10 strains revealed the robustness with only 4 false positive (out of 950 measurements) reads. The complete MIC information for the ten clinical isolates obtained using the inventive system to perform what may be designated in-fiber antibiotic susceptibility testing (FAST) is provided in Table 1. There were some entries that disagreed with the clinical data. For example, it was found the MIC for ampicillin for the CIET-10 strain as >16 pg/ml where the clinical report suggested that it was £16 pg/ml. The breakpoint (BP) value for ampicillin for E. Co// is 8 pg/ml. Since the AST panel that was used cannot determine the exact MIC for ampicillin at this point, the discrepancy for this case might be considered as minor as both tests indicate some resistance against ampicillin.

In five other cases where there are discrepancies between the NM-43 AST panel and FAST, the errors for the cases of Cefotaxime (BP = 0.25 pg/ml, FAST MIC > 32 pg/ml, and NM-43 MIC £16 pg/ml ), Cefoxitin (BP 4 pg/ml, FAST MIC£4 pg/ml

NM-43 MIC > 16 pg/ml), Ceftadizime (BP 0.25 pg/ml, FAST MIC > 16 pg/ml, NM-43

MIC > 8 pg/ml (CIET-5), £1 pg/ml (CIET-9) could be classified as major errors. In the last case of Ertapenem, an MIC £ 1 was found versus £ 2 by the plate method, however the EUCAST system does not indicate a breakpoint value for this drug. These differences might be due to experimental error or greater sensitivity of the inventive system. Finally, none of the measurements yielded false negative (FN) measurements where growth of a bacterial strain was missed in a drug condition where it was expected to grow. These findings further articulate the power of the inventive system and the FAST assay.

In the present disclosure, a novel optofluidic system has been described that has up to 4 orders of magnitude enhanced sensitivity compared to conventional optical methods. The detection system platform is a fast, highly sensitive, and low-cost optical bacterial detection system that can be used as a field-deployable device. The system can detect bacterial growth in less than an hour even at room temperature and quantitatively measure antibiotic sensitivity of pathogenic bacteria within two hours.

The platform is open to further improvements with implementing further techniques, such as integration of advanced microfluidic and particle separation systems, antibody-antigen interacting surfaces, photonic-crystal fibers, and improved electronics, which can elevate it to a fully capable point of care direct Separation- detection-identification instrument.

Compared with existing detection methods, the disclosed detection method provides a few key advantages in sensitive and rapid detection: The detection can be performed in both low- and high-volume samples by engineering the hollow tube volume. Higher volumes for a constant tube diameter helps the detection limit improve since it is not limited by the volume of the liquid but by the cross-section of the measurement chamber. While the current state of the art technologies have assay times up to several hours, the disclosed detection system provides instant detection, meaning the measurement is performed spontaneously for the encapsulated volume within the hollow tube. Thus, by engineering the length and diameter of the tube, the detection system has the capability of adjustable sensitivity level, pushing the single cell limit.