We demonstrate a novel optical method for the detection and differentiation between erythrocytes and leukocytes that uses amplitude and phase information provided by optical coherence tomography (OCT). Biological cells can introduce significant phase modulation with substantial scattering anisotropy and dominant forward-scattered light. Such physical properties may favor the use of a trans-illumination imaging technique. However, an epi-illumination mode may be more practical and robust in many applications. This study describes a new way of measuring the phase modulation introduced by flowing microobjects. The novel part of this invention is that it uses the backscattered signal from the substrate located below the flowing/moving objects. The identification of cells is based on phase-sensitive OCT signals. To differentiate single cells, a custom-designed microfluidic device with a highly scattering substrate is introduced. The microchannels are molded in polydimethylsiloxane (PDMS) mixed with titanium dioxide (TiO2) to ensure high scattering properties. The statistical parameters of the measured signal depend on the cells’ features, such as their size, shape, and internal structure.
© 2015 Optical Society of America
One of the largest challenges in modern medicine is the ability to perform precise and reliable diagnostic laboratory tests in developing countries and remote areas that lack conventional analytical laboratories. There is a real need for a compact, robust, and easy-to-operate class of medical devices that enables low-cost, accurate, and rapid point-of-care diagnostic laboratory tests without requiring the infrastructure of a complete analytical laboratory [1, 2]. The direct identification of cells forms the basis of many analytical laboratory tests including those used in hematology and oncology diagnostics. One of the most important examples is the complete blood count (CBC), which is the most widely used clinical test ordered for almost every disease. A foundation of the CBC test is the use of specialized hematology analyzers based on flow cytometry  or the Coulter  detection method to count blood cells.
Most existing cell counting methods are based on optical detection techniques due to their minimal invasiveness, high registration speed, high spatial resolution, and ability to detect many physical parameters including scattering, polarization, absorption, and fluorescence, etc. The clinical and biomedical research applications of optical flow cytometry comprise various cell counting, cell-based fluorescent immunoassays, biomarker detection, and DNA stains, among many others . However, the most commonly used flow cytometers and automated cell counters are sizable, complex, and usually expensive. Moreover, such devices require the use of chemical reagents. The sample preparation process may be complicated and may need to be performed by professional medical personnel . Over the last few years, new methods for optical cytometry that employ modern optical imaging have been developed to address some of these issues. A few interesting non-microfluidic optical methods have been proposed, such as diffraction imaging flow cytometry [7, 8], in vivo flow cytometry based on two-photon autofluorescence imaging [9, 10], label-free spectrally encoded flow cytometry , (lens-free) holographic microscopy [12, 13], cell-phone based fluorescent microscopy [14, 15], diffraction phase microscopy , quantitative absorption cytometry , and optical coulter counter .
One of the most promising optical detection technologies that has been developed is optical coherence tomography (OCT). OCT is a powerful, high-resolution, and high-speed optical imaging modality that is non-contact and non-invasive. It can be used to obtain three-dimensional reconstructions of weakly scattering biological tissues with moderate spatial resolution but rather large imaging depth . One of the most important advantages of OCT is its ability to provide both the phase and the intensity of backscattered light in a single measurement . Another good point of this technology is its capability of performing in vivo imaging; thus, OCT is currently a most promising medical modality. OCT has been previously applied to the measurement and imaging of flow in microchannels, e.g., imaging of electro-osmotic flow (EOF) dynamics , secondary flow and two-liquid mixing , and capillary-driven blood flow . It is also used for measurements of Intralipid [24, 25] and human blood  flow velocity profiles. Such measurements in microchannels are beneficial to sample volume minimization, fast sequential sample scanning, and rapid sample content detection. However, to the best of our knowledge, OCT has not yet been applied to blood cell count and differentiation in microfluidic microchannel systems.
Recently, lab-on-a-chip platforms combining microfluidics  with optical methods  have attracted significant attention because of their potential use in blood testing [29–31] and in counting and discriminating various kinds of biological cells, including leukocyte (differential) counting [32–34] and classification , tumor cell counting [35, 36] and identification , and T CD4 + cell counting [38, 39], among others. Most of these devices can be considered miniaturized lab-on-a-chip flow cytometers that employ optofluidics [6, 40], color-space-time coding , microfluidic compact discs , Raman spectroscopy with optical traps , optical force measurements , optical tweezers , fluorescence detection , and lens-free holographic microscopy , among others. Some of these devices are promising for implementation in compact, simple, and inexpensive point-of-care or personal-use medical equipment. However, most of the existing microfluidic optical flow cytometry methods require complicated sample preparation processes, such as (antibody) labeling [32, 33] or staining to detect or separate specific populations or subpopulations of cells from the analyzed sample [35, 36]. In addition, almost all existing flow cytometry techniques, including commonly used conventional flow cytometry devices, are based only on the information encoded in the intensity of light that is scattered by the measured object.
In this paper we present a novel optical method that enables to detect and differentiate moving microobjects like erythrocytes and leukocytes. This new method uses a phase-sensitive OCT system with dedicated microfluidic setup. The significant difference from standard microfluidic systems is that the microchannels are molded in polydimethylsiloxane (PDMS) mixed with titanium dioxide (TiO2) to ensure high scattering properties of the substrate. This modification is easy and inexpensive to apply in the fabrication process. We also conduct a case study in which we optically identify and differentiate flowing blood cells. According to our knowledge, the main advancement in our approach is that objects are identified based on OCT signals coming from a stationary scattering medium localized beneath the flowing/moving object (EP Application No. 14166012.6); these signals are stronger than those directly scattered from the object. The statistical parameters of these signals depend on the features of the object, including its size, shape, and internal structure . This effort can be also considered to be a first step toward non-invasive, in vivo detection and recognition of blood cells directly within blood vessels, which is a Holy Grail of modern hematology and medical diagnostics.
A number of methods can be used to differentiate microobjects (cells, bacteria, etc.) on the basis of light scattered directly from the objects, for example, conventional optical flow cytometry methods . Our approach is based on optical coherence tomography, which reconstructs information regarding the depth-resolved structure of the sample by using optical interferometry. OCT has two main versions: (1) time-domain OCT (Td-OCT), first proposed by Huang et al. ; and (2) Fourier-domain OCT (Fd-OCT). The latter considerably improves the imaging speed and detection sensitivity compared to Td-OCT [45, 46]. The Fd-OCT setup, which comprising an interferometer, a spectrometer, and a light source (or swept light source), can also be greatly miniaturized [47–50].
Ophthalmic in vivo imaging remains the most successful application of clinical OCT [51, 52]. The initial idea for our new method is based on OCT retinal imaging, presented in Fig. 1. This figure illustrates the significance of the phase modulation introduced by flowing blood. Such modulation is manifested by characteristic intensity shadows that are localized just below the blood vessels, which are barely visible in a standard OCT cross-sectional image based on a backscattered intensity image, presented in Fig. 1(a). However, the same modulation is clearly visualized using the phase-sensitive mode of OCT imaging , as shown in Figs. 1(b)–1(c). The cross section of the blood vessels and the region localized below the vessels show a strong variation of the phase value in the phase-sensitive image. These observations motivated us to utilize this strong effect and to develop a new method of blood cell differentiation using both the intensity and phase information from OCT signals. Based on these observations, we introduce a setup that combines Fd-OCT with a specially designed microfluidic device that in some sense imitates blood vessels located on the retina of the human eye. The microchip’s microchannels emulate retina microcapillaries, while the scattering substrate corresponds to retina tissue (the scattering medium).
In this way, we attempt to differentiate objects using light that is initially disturbed by the object and secondarily scattered by a material of known optical properties. The idea is presented in Fig. 2; in our approach the backscattered light is registered by low-coherence interferometry, in particular by Fd-OCT. In the configuration presented in Fig. 2(a), the phase modulation introduced by the object flowing in a microchannel will also contribute to the electromagnetic field scattered later from the base. For a stationary location of an illuminatingbeam (M-scan mode) and with no object flowing in the channel, the registered OCT signal from the scattering base will be static. An additional contribution to the phase modulation induced by the object will be clearly visible on this static background as a new speckle field. This effect is presented in Fig. 2(b), and will be observable in both the intensity and the phase-sensitive images. Therefore, the scattering medium can be considered to be a kind of “intensifier” of the phase modulation introduced by the object. In order to quantify and parameterize this information, rectangular regions of interest (ROI) are selected in the area of the scattering base spanning horizontally over the entire modulation signal.
The selected data are then used for calculating a set of statistical parameters which depend on the features of the object (size, shape, internal structure, etc.). Therefore, this method can be used for the optical differentiation of every type of small object, such as cancer cells, nerve cells, beads, biological and artificial grains, etc., as long as they differ in size, shape, or internal structure and thus influence the light scattering process.
2.2 Optical system
The complete experimental instrumentation used in this work is presented in Fig. 3(a). The system combines an optical setup (Fd-OCT) with a microfluidic device. The microfluidic system includes a microfluidic chip, syringe pump (AP22, Ascor), a syringe with a measured sample, polythene hoses (0.76 mm in diameter), and a biohazard waste/sharps container. The syringe pump is used to control the flow of the diluted sample (blood cells) inside the microfluidic setup, which consists of a coverslip and microchannels formed in the top substrate layer (scattering medium) attached to the coverslip. The backscattered light coming from the measured region of the microfluidic chip is collected by the same optical system and then registered.
In these experiments, we separately studied the flowing blood cells—i.e., erythrocytes and leukocytes—as two separate samples. For these experiments a light beam is fixed and focused on one of the microchannels in the microfluidic chip. During the time of measurement the cells are scanned within the light beam as long as the flow is sustaining by the syringe pump (flow through the microchannel). This type of scan in OCT is also called an M-scan.
For this study, we designed and constructed a phase-sensitive Fd-OCT optical setup as the core of the experimental system depicted in Fig. 3. Figure 4(a) presents the schematic of the experimental OCT setup. It includes a fiber coupler (70/30 beam splitter with four optical fiber endings, AC Photonics); a dedicated spectrometer with high-speed line scan camera (sp4096-140k, Sprint, Basler; 2048 pixels used in the measurements); dispersion compensation elements (prisms); a mirror in the reference arm; and two options for the objective lens: a 10X OCT Scan Lens (LSM02-BB, Thorlabs) for the OCT configuration or, alternatively, 20X, 0.45 N.A., W.D. 6.6–7.8 mm, (LUCPLFLN 20X, Olympus). Table 1 shows the setup parameter values depending on the applied optical configuration. The light source is a broadbandfemtosecond laser (FemtoSource Fusion, Femtolasers) with center wavelength λ0 = 790 nm and FWHM = 150 nm. This setup also includes optical fibers and several lenses for controlling the light beam parameters inside the reference and object arms of the interferometer. The optical setup is connected to the computer via a frame grabber PCI Express card (PCIe-1433, National Instruments), and C + +/LabVIEW coded software is used to read out and process the measurement data.
We measure M-scans instead of the classical B-scans (cross-sectional images of the scanned object). This type of measurement was intentionally designed as part of an invented method to obtain static background signals coming from an optically uniform scattering medium.
A single OCT measurement includes 100 intensity and 100 phase (phase change) M-scans, with 2,000 A-scans (single axial depth scan) per one M-scan. Differential intensity M-scans are automatically calculated from the intensity M-scans by subtracting the constant background signal coming from the static scattering base placed beneath the microchannels. This operation is depicted in Fig. 4(b). Thus, only information coming from moving objects should be visible in the differential M-scan image. In order to obtain these results—i.e., M-scan images with registered signals derived from flowing objects—6 μs exposition time and 8 μs repetition time were set on the spectrometer camera (Table 1) and 3 ml/h flow speed was set on the syringe pump. These three values were selected experimentally to achieve satisfactory quality of the results (see Results and discussion).
2.3 Microfluidic chip
The microfluidic chip was fabricated using rapid prototyping . The microchannels were molded in the PDMS substrate mixed with 1 wt% titanium white (titanium dioxide, TiO2). The titanium white was used as an additional constituent to ensure the strong scattering property of the substrate so it can be used as a scattering medium in the microfluidic sample setup (see Methods).
The TiO2-PDMS channels were bonded to a glass microscope coverslip after high RF oxygen-plasma exposure. The microfluidic chip comprises four sections of the microchannel structure. One section contains the sample inlet and outlet (both located on the substrate underside) and their connecting channels (300 μm wide and 40 μm high). These channels branch into three or six smaller microchannels, depending on the section of the microchannel structure (two sections with three microchannels and two sections with six microchannels). We used both kinds of microchannels in our experiments. Figure 5(a) presents a schematic diagram of the microfluidic chip comprising one of four normally existing sections of an entire microchannel structure. The inlet and outlet channels split into three microchannels (in Fig. 5(a), n = 3, a = 300 μm, b = 100 μm, and h = 40 μm). Figure 5(c) depicts a microscopic image of a fragment of the microchannel structure with a six-microchannel section (n = 6, a = 300 μm, b = ∼50 μm, and h = 40 μm). Figure 5(d) reveals an image of the inlet/outlet surroundings of one section of the microchannel structure.
2.4 Blood samples and peripheral blood leucocytes isolation
Peripheral blood was obtained from healthy volunteers (blood donors, median age 30 years, range 20-35). The material was collected after receiving approval from the Bioethical Committee of the Collegium Medicum in Bydgoszcz, Poland (according to the Helsinki declaration). Each volunteer was familiarized with the objectives of the study and expressed written consent for material collection.
Venous blood samples (9 ml) from healthy volunteers were drawn into tubes containing heparin as anticoagulant (Vacuette, Greiner Bio-One). In order to validate the method of differentiating the microobjects, two separate samples of blood cells were characterized, namely erythrocytes (RBC) and leukocytes (WBC). To prepare the RBCs, whole blood was dissolved in Alsever's solution (Sigma-Aldrich) to obtain a 1% solution. We decided to use diluted blood as the RBC sample due to very high ratio of RBCs to WBCs in whole blood (order of 1000:1). In order to prepare the WBC sample, the following procedure was performed. First, anti-coagulated whole blood was diluted 1:1 with a lysing solution (Becton Dickinson Bioscience) in sterile, non-adherent, 50 ml polypropylene tubes (Falcon, Corning). Then, the specimens were incubated for 15–20 minutes in the dark at room temperature and centrifuged at 300 × g for 10 minutes at 4°C. Next, the cells were washed three times by adding 50 ml of cold (2–8°C), sterile buffered saline (PBS, PAA). They were centrifuged at 300 × g for 10 minutes at 4°C. The specimens of isolated peripheral blood leucocytes were analyzed with the FACScan flow cytometer (Becton Dickinson Bioscience). A total of 50,000 events were collected. All flow cytometry data analyses were performed using FlowJo software ver. 7.6.1. (TreeStar). The representative frequencies of the major cell populations of granulocytes, monocytes, and lymphocytes are shown in Fig. 6. The purity of the isolated leukocytes was greater than 96%. Next, the isolated cells were resuspended in PBS (PAA) and subjected to further analysis.
2.5 Data analysis
In order to differentiate erythrocytes from leukocytes, the 1st- and 2nd-order speckle statistics of the acquired sections of data are considered. We identified four of the most effective statistical parameters out of 24 proposed initially; two of them are intensity-based (vertical (axial) speckle size and standard deviations (intensity)) and two are phase-based (standard deviations (phase) and contrast top-middle).
The estimation of the mean speckle size (2nd-order statistics) relies on a statistical analysis of the speckle intensity distributions based on the Power Spectral Density (PSD) and the autocorrelation function of these distributions. The PSD of a signal is defined by the module square of its Fourier transform (). A spatial distribution of the speckle intensity is represented by a 2D matrix, , which corresponds to the ROI of M-scans data. Then, the PSD of the intensity distribution is defined as55, 56]. Taking into account the measurement geometry of the OCT M-scans, the horizontal speckle size corresponds to the direction parallel to the flow inside the microchannel, while the vertical direction of the speckle corresponds to the axial OCT direction parallel to the focused light beam.
The parameter standard deviations is calculated identically for both types of M-scan ROIs (intensity and phase change). For each ROI, a special window is applied in order to calculate the standard deviation, so that the window is two pixels wide and its height is equal to the height of the ROI in pixels. The constructed window moves pixel by pixel until it achieves the right boundary of the ROI, and the standard deviation is calculated simultaneously from values covered by the window for each position. Finally, we average the values of the standard deviation obtained in this manner for each window position. Let us assume that the image of the M-scan ROI has a width M and height N (in pixels). If we designate the value at the position i inside the window placed on the k position within the section image as , then the expression of the standard deviations parameter Γ is:
Contrast top-middle is an unique phase-based parameter calculated in the spatial frequency domain (Fourier domain). The magnitude of the 2D Fourier transform obtained for selected M-scan ROI data is calculated first: , where indicates the 2D Fourier transform, is the real part of , and represents the imaginary part. The value of the parameter is obtained by dividing the summarized values coming from two specially selected regions (windows) of the magnitude data. This is the measure of the contrast between the mean intensity levels of these regions. Figure 7(a) shows how the contrast top-middle parameter is constructed with two regions indicated (“top” and “middle”).
The automatic differentiation of tens of thousands of cells (∼70,000) from one set of measurements (more than 100 GB of data) is made possible by using custom-designed, automated LabVIEW software with an implemented algorithm for cell differentiation. Figure 7(b) depicts a flowchart of the complete algorithm. The program loads consecutive OCT measurement files (OCT files), and for each one calculates 100 intensity M-scans consisting of 2,000 A-scans. A full A-scan consists of 8,192 pixels after implementing the zero-padding operation (4 × 2048 active camera pixels). For further analysis, only 700-pixel-long parts of full A-scans are calculated comprising the ROIs (modulation signals). This allows us to obtain intensity M-scans that are 700 pixels high and 2000 pixels wide and constitute the base for calculating phase M-scans. This operation is required to perform the automatic localization of modulation signals (described below), and it significantly contributes to the efficiency of the calculations by avoiding data redundancy. The differential intensity M-scans are calculated for an intensity-based analysis (vertical speckle size and standard deviations for intensity). Figure 7(c) depicts an automatic detection algorithm and spatial localization of the modulation signals (intensity and phase) by calculating their coordinates , where i defines the consecutive modulation signals detected in both types of M-scans. The algorithm for the detection and localization of the modulation signals is based on a numerical analysis of axial (differential intensity) and transversal (intensity) cumulative M-scans characteristics. The Xi coordinates computed from the differential intensity M-scans are also used to localize the corresponding signals in phase M-scans, while the Y coordinate is calculated once from the first analyzed intensity M-scan of a given OCT file. Then, the aforementioned ROIs (rectangular sections of data) are selected automatically for each localized signal to calculate the statistical parameters for cell differentiation.
3. Results and discussion
In order to demonstrate the applicability of the method for distinguishing between microobjects, we conducted an experiment to differentiate erythrocytes (RBC) from leukocytes (WBC). It is well known that WBC and RBC have diverse scattering properties due to their differences in size and internal structure . Generally, erythrocytes (∼7.5 μm in diameter) are smaller than leukocytes (the diameter ranges from a minimum of 8 μm for lymphocytes up to maximum of 25 μm for monocytes)  and they have no nucleus. Therefore, we expect to see a qualitative difference between these cells by observing the modulation signals in M-scan images and a quantitative difference by calculating the statistical differential parameters.
The OCT setup configuration with a 20X objective lens was used to obtain the results presented in Fig. 8. Some of the setup parameters are listed in Table 1. During the experiment, the detection sensitivity of the setup was approximately 95 dB, the power of light directed to the sample was 470 μW, and the imaging depth was ∼2 mm. A previously prepared blood sample (diluted blood or isolated WBC solution) was drawn into a plastic syringe and placed in the syringe pump. Next, the cells were pumped into the microchannel (microfluidic sample setup) with a flow rate of 3 ml/h. Because of scattering medium inhomogeneity, the microfluidic chip was fixed spatially to ensure that all measurements were taken exactly in the same place. To avoid data redundancy, we computed 700 × 2000-pixel parts (700 × 1998 pixels for phase data) of full 2D M-scans data (8192 × 2000 / 8192 × 1998 pixels after the zero-padding operation) comprising the information of interest. We set M-scans ROI sizes to 28 × 120 pixels for the intensity data and 28 × 75 pixels for the phase data for both the RBC and WBC samples. These 28-pixel-wide windows covered the entire modulation signals transversally. The intensity ROI was set to be longer (120 pixels) than the phase ROI (75 pixels) due to the specificity of the vertical speckle size calculations. Moreover, all ROIs were selected at the same distance (in pixels) from the boundary between the microchannel and the substrate.
Figure 8(a) shows examples of two different modulation signals coming from two different RBCs (diluted whole blood), depicted separately in both differential intensity M-scan and phase M-scan images. Both types of images are also shown in Fig. 8(b) to depict the two different modulation signals from WBCs. Figure 8(c) shows two enlarged sections (28 × 75 pixels) of RBC and WBC phase change modulation signal images (phase M-scan) from Figs. 8(a)–8(b), respectively. The modulation signals corresponding to both types of cells (RBC and WBC) are considerably different and can be distinguished qualitatively; i.e., RBCs phase values are more randomized in comparison to WBCs. Furthermore, WBCs are characterized by bigger and brighter speckles comparing to RBCs, as it is presented in enlarged rectangular sections (ROIs) of the intensity images presented in Figs. 8(a)–8(b).
Different sets and combinations of the four most effective statistical differential parameters are depicted in Figs. 8(d)–8(j), as 1D, 2D, and 3D scatter plots including both RBC and WBC sample data. The values of these parameters were obtained by a numerical analysis of single modulation signals derived from corresponding flowing cells, i.e., the single parameter value corresponds to single cell flows in the microchannel (RBC or WBC). In all plots shown in Figs. 8(d)–8(j), red dots (points) indicate parameter values related to the RBC sample, while black dots indicate the WBCs. Figures 8(d)–8(f) reveal one-dimensional comparisons of one parameter values in the order that consecutive signals were registered during the measurement for RBC and WBC samples separately. Multidimensional scatter plots presented in Figs. 8(g)–8(j) constitute a combination of two (2D) or three (3D) statistical parameter histograms.
All quantitative results shown in Figs. 8(d)–8(j) reveal that the statistical parameters used for the analysis enable satisfactory separation of erythrocytes and leukocytes. Moreover, the 2D scatter plot presented in Fig. 8(i) proves that we are able to differentiate between cells considering only the phase OCT information (two phase-based differential parameters). As we expected, the most satisfactory separation is provided by a 3D scatter plot presented in Fig. 8(j), comprising the combination of three statistical parameters.
We propose and describe a novel optical method for the detection and differentiation of microobjects (on the order of micrometers) that uses amplitude and phase information provided by OCT. It was demonstrated that this method can be used successfully to detect and differentiate two morphotic elements present in human blood, i.e., erythrocytes and leukocytes. Unlike conventional optical flow cytometry, our method provides additional quantitative information at the single cell level as a result of a statistical analysis of the phase modulation introduced by flowing cells. In other words, this technique enables single-cell statistics (additional information at the single-cell level), instead of statistics of all measured cells as with conventional cytometry methods [3, 5].
Based on this experimental and theoretical work, we conclude that OCT phase information is very promising for the effective differentiation of objects flowing in microfluidic devices. Therefore, improved phase stability of a phase-sensitive OCT device can be introduced by careful optical optimization of the optical setup in the so-called common-path configuration. In this contribution, we determined the optimal experimental conditions required to obtain satisfactory results such as high OCT detection sensitivity, the proper acquisition parameter values to set on the camera (Table 1), and the required width range of the registered modulation signals (provided in pixels) for the corresponding flow rate value set on the syringe pump.
Far-reaching goal of our studies is in vivo imaging. Before we reach this point, we have to perform experiments on whole blood. Currently, we are not able to perform such experiments since the ratio of erythrocytes to leukocytes is on the order of 1000:1. Assuming ultrafast online calculations, the acquisition process only would take approximately 14 hours to analyze 100 times more cells (∼7,000,000 cells with ∼7,000 WBCs statistically) in whole blood (∼3,200,000 M-scans acquired with a single M-scan measured in 16 ms time using Basler Sprint sp4096-140k spectrometer camera). To address this issue and conduct measurements for whole blood, we plan to introduce the smart dosing of blood cells in a microfluidic device (hydrodynamic focusing, for instance) in order to minimize the number of “void” measurements without phase modulation (many more cells analyzed in a single M-scan). Additionally, we plan to optimize the software by its implementation on graphics cards (CUDA) in order to considerably increase the efficiency of the calculations.
It is important to mention that current setup can be simplified and optimized in future. We used femtosecond laser in order to control the experiment and visualize the single blood cells flowing in the microfluidic channel. This step can be skipped in the default instrument since we are interested only in the signals coming from the scattering medium. Therefore, the cost of the default system may be dramatically decreased by reducing the axial resolution and using SLD as a light source. Optical components and compact spectrometers dedicated to OCT also became less expensive since OCT is a standard ophthalmic diagnostic tool selling yearly in thousands of units.
We also believe that, aside from blood cells, the proposed method can be used for the optical detection and differentiation of other types of small objects, such as cancer cells, beads, biological and artificial grains, etc., providing they differ in size, shape, or internal structure. Potentially this method can be also utilized in the chemical and pharmaceutical industry for cleanliness inspections during the production of drugs or chemicals, as a means of testing their quality or possible contamination.
Paweł Ossowski and Maciej Wojtkowski acknowledge support from the National Science Center within the MAESTRO Programme (DEC-2011/02/A/ST2/00302). The authors thank Iwona Gorczyńska, PhD, Szymon Tamborski, MSc, Janusz Strzelecki, PhD, and Maciej Szkulmowski, PhD, for technical help.
References and links
1. P. Yager, T. Edwards, E. Fu, K. Helton, K. Nelson, M. R. Tam, and B. H. Weigl, “Microfluidic diagnostic technologies for global public health,” Nature 442(7101), 412–418 (2006). [CrossRef] [PubMed]
2. W. Jung, J. Han, J.-W. Choi, and C. H. Ahn, “Point-of-care testing (POCT) diagnostic systems using microfluidic lab-on-a-chip technologies,” Microelectron. Eng. 132, 46–57 (2015). [CrossRef]
3. H. M. Shapiro, Practical flow cytometry (John Wiley & Sons, 2005).
4. W. H. Coulter, “High speed automatic blood cell counter and cell size analyzer,” Proc. Natl. Electron. Conf.12, 1034–1040 (1956).
5. V. V. Tuchin, Advanced Optical Flow Cytometry: Methods and Disease Diagnoses (John Wiley & Sons, 2011).
8. K. Dong, Y. Feng, K. M. Jacobs, J. Q. Lu, R. S. Brock, L. V. Yang, F. E. Bertrand, M. A. Farwell, and X.-H. Hu, “Label-free classification of cultured cells through diffraction imaging,” Biomed. Opt. Express 2(6), 1717–1726 (2011). [CrossRef] [PubMed]
9. Y. Zeng, J. Xu, D. Li, L. Li, Z. Wen, and J. Y. Qu, “Label-free in vivo flow cytometry in zebrafish using two-photon autofluorescence imaging,” Opt. Lett. 37(13), 2490–2492 (2012). [CrossRef] [PubMed]
10. Y. Zeng, J. Xu, D. Li, L. Li, Z. Wen, and J. Y. Qu, “In vivo cytometry using two-photon autofluorescence microscopy,” in Biomedical Optics and 3-D Imaging, OSA Technical Digest, (Optical Society of America, 2012), BSu3A. 20.
11. L. Golan, D. Yeheskely-Hayon, L. Minai, E. J. Dann, and D. Yelin, “Noninvasive imaging of flowing blood cells using label-free spectrally encoded flow cytometry,” Biomed. Opt. Express 3(6), 1455–1464 (2012). [CrossRef] [PubMed]
12. S. Seo, S. O. Isikman, I. Sencan, O. Mudanyali, T.-W. Su, W. Bishara, A. Erlinger, and A. Ozcan, “High-throughput lens-free blood analysis on a chip,” Anal. Chem. 82(11), 4621–4627 (2010). [CrossRef] [PubMed]
13. B. Rappaz, A. Barbul, Y. Emery, R. Korenstein, C. Depeursinge, P. J. Magistretti, and P. Marquet, “Comparative study of human erythrocytes by digital holographic microscopy, confocal microscopy, and impedance volume analyzer,” Cytometry A 73(10), 895–903 (2008). [CrossRef] [PubMed]
15. H. C. Koydemir, Z. Gorocs, D. Tseng, B. Cortazar, S. Feng, R. Y. L. Chan, J. Burbano, E. McLeod, and A. Ozcan, “Rapid imaging, detection and quantification of Giardia lamblia cysts using mobile-phone based fluorescent microscopy and machine learning,” Lab Chip 15(5), 1284–1293 (2015). [CrossRef] [PubMed]
17. E. Schonbrun, R. Malka, G. Di Caprio, D. Schaak, and J. M. Higgins, “Quantitative absorption cytometry for measuring red blood cell hemoglobin mass and volume,” Cytometry A 85(4), 332–338 (2014). [CrossRef] [PubMed]
18. E. Schonbrun and G. Di Caprio, “Differentiating neutrophils using the optical coulter counter,” in Biomedical Optics (BiOS)2015, (SPIE, 2015), 933614–1-6.
19. D. Huang, E. A. Swanson, C. P. Lin, J. S. Schuman, W. G. Stinson, W. Chang, M. R. Hee, T. Flotte, K. Gregory, C. A. Puliafito, and et, “Optical coherence tomography,” Science 254(5035), 1178–1181 (1991). [CrossRef] [PubMed]
20. M. Szkulmowski, A. Szkulmowska, T. Bajraszewski, A. Kowalczyk, and M. Wojtkowski, “Flow velocity estimation using joint Spectral and Time domain Optical Coherence Tomography,” Opt. Express 16(9), 6008–6025 (2008). [CrossRef] [PubMed]
21. L. Wang, W. Xu, M. Bachman, G. Li, and Z. Chen, “Phase-resolved optical Doppler tomography for imaging flow dynamics in microfluidic channels,” Appl. Phys. Lett. 85(10), 1855–1857 (2004). [CrossRef]
23. S. Cito, Y.-C. Ahn, J. Pallares, R. M. Duarte, Z. Chen, M. Madou, and I. Katakis, “Visualization and measurement of capillary-driven blood flow using spectral domain optical coherence tomography,” Microfluid. Nanofluidics 13(2), 227–237 (2012). [CrossRef] [PubMed]
24. J. Lauri, J. Czajkowski, R. Myllylä, and T. Fabritius, “Measuring flow dynamics in a microfluidic chip using optical coherence tomography with 1µm axial resolution,” Flow Meas. Instrum. 43, 1–5 (2015). [CrossRef]
25. J. Lauri, M. Wang, M. Kinnunen, and R. Myllylä, “Measurement of microfluidic flow velocity profile with two Doppler optical coherence tomography systems,” in Biomedical Optics (BiOS)2008, (SPIE, 2008), 68630F–68630F–68638.
26. D. M. Bukowska, L. Derzsi, S. Tamborski, M. Szkulmowski, P. Garstecki, and M. Wojtkowski, “Assessment of the flow velocity of blood cells in a microfluidic device using joint spectral and time domain optical coherence tomography,” Opt. Express 21(20), 24025–24038 (2013). [CrossRef] [PubMed]
27. J. Godin, C. H. Chen, S. H. Cho, W. Qiao, F. Tsai, and Y. H. Lo, “Microfluidics and photonics for Bio-System-on-a-Chip: A review of advancements in technology towards a microfluidic flow cytometry chip,” J. Biophotonics 1(5), 355–376 (2008). [CrossRef] [PubMed]
31. A. Kummrow, J. Theisen, M. Frankowski, A. Tuchscheerer, H. Yildirim, K. Brattke, M. Schmidt, and J. Neukammer, “Microfluidic structures for flow cytometric analysis of hydrodynamically focussed blood cells fabricated by ultraprecision micromachining,” Lab Chip 9(7), 972–981 (2009). [CrossRef] [PubMed]
32. H. Yun, H. Bang, J. Min, C. Chung, J. K. Chang, and D.-C. Han, “Simultaneous counting of two subsets of leukocytes using fluorescent silica nanoparticles in a sheathless microchip flow cytometer,” Lab Chip 10(23), 3243–3254 (2010). [CrossRef] [PubMed]
33. W. Shi, L. Guo, H. Kasdan, and Y.-C. Tai, “Four-part leukocyte differential count based on sheathless microflow cytometer and fluorescent dye assay,” Lab Chip 13(7), 1257–1265 (2013). [CrossRef] [PubMed]
34. D. Vercruysse, A. Dusa, R. Stahl, G. Vanmeerbeeck, K. de Wijs, C. Liu, D. Prodanov, P. Peumans, and L. Lagae, “Three-part differential of unlabeled leukocytes with a compact lens-free imaging flow cytometer,” Lab Chip 15(4), 1123–1132 (2015). [CrossRef] [PubMed]
35. M. Zhao, P. G. Schiro, J. S. Kuo, K. M. Koehler, D. E. Sabath, V. Popov, Q. Feng, and D. T. Chiu, “An automated high-throughput counting method for screening circulating tumor cells in peripheral blood,” Anal. Chem. 85(4), 2465–2471 (2013). [CrossRef] [PubMed]
36. L. S. Lim, M. Hu, M. C. Huang, W. C. Cheong, A. T. L. Gan, X. L. Looi, S. M. Leong, E. S.-C. Koay, and M.-H. Li, “Microsieve lab-chip device for rapid enumeration and fluorescence in situ hybridization of circulating tumor cells,” Lab Chip 12(21), 4388–4396 (2012). [CrossRef] [PubMed]
37. S. Dochow, C. Krafft, U. Neugebauer, T. Bocklitz, T. Henkel, G. Mayer, J. Albert, and J. Popp, “Tumour cell identification by means of Raman spectroscopy in combination with optical traps and microfluidic environments,” Lab Chip 11(8), 1484–1490 (2011). [CrossRef] [PubMed]
38. X. Cheng, D. Irimia, M. Dixon, K. Sekine, U. Demirci, L. Zamir, R. G. Tompkins, W. Rodriguez, and M. Toner, “A microfluidic device for practical label-free CD4+ T cell counting of HIV-infected subjects,” Lab Chip 7(2), 170–178 (2007). [CrossRef] [PubMed]
39. Z. Wang, S. Y. Chin, C. D. Chin, J. Sarik, M. Harper, J. Justman, and S. K. Sia, “Microfluidic CD4+ T-cell counting device using chemiluminescence-based detection,” Anal. Chem. 82(1), 36–40 (2010). [CrossRef] [PubMed]
40. D. Wu, J. Xu, L.-G. Niu, S.-Z. Wu, K. Midorikawa, and K. Sugioka, “In-channel integration of designable microoptical devices using flat scaffold-supported femtosecond-laser microfabrication for coupling-free optofluidic cell counting,” Light Sci. Appl. 4(1), e228 (2015). [CrossRef]
41. S. M. Imaad, N. Lord, G. Kulsharova, and G. L. Liu, “Microparticle and cell counting with digital microfluidic compact disc using standard CD drive,” Lab Chip 11(8), 1448–1456 (2011). [CrossRef] [PubMed]
43. J. Martini, M. I. Recht, M. Huck, M. W. Bern, N. M. Johnson, and P. Kiesel, “Time encoded multicolor fluorescence detection in a microfluidic flow cytometer,” Lab Chip 12(23), 5057–5062 (2012). [CrossRef] [PubMed]
44. A. Hoekstra, V. Maltsev, and G. Videen, Optics of Biological Particles:Proceedings of the NATO Advanced Research Workshop on Fluorescence and Other Optical Properties of Biological Particles for Biological Warfare Agent Sensors,Novosibirsk, Russia,3–6 October 2005 (Springer Science & Business Media, 2007), Vol. 238.
46. M. Wojtkowski, A. Kowalczyk, R. Leitgeb, and A. F. Fercher, “Full range complex spectral optical coherence tomography technique in eye imaging,” Opt. Lett. 27(16), 1415–1417 (2002). [CrossRef] [PubMed]
47. G. Yurtsever, B. Považay, A. Alex, B. Zabihian, W. Drexler, and R. Baets, “Photonic integrated Mach-Zehnder interferometer with an on-chip reference arm for optical coherence tomography,” Biomed. Opt. Express 5(4), 1050–1061 (2014). [CrossRef] [PubMed]
48. P. Pottier, M. J. Strain, and M. Packirisamy, “Integrated Microspectrometer with Elliptical Bragg Mirror Enhanced Diffraction Grating on Silicon on Insulator,” ACS Photonics 1(5), 430–436 (2014). [CrossRef]
49. G. Yurtsever, N. Weiss, J. Kalkman, T. G. van Leeuwen, and R. Baets, “Ultra-compact silicon photonic integrated interferometer for swept-source optical coherence tomography,” Opt. Lett. 39(17), 5228–5231 (2014). [CrossRef] [PubMed]
50. S. Schneider, M. Lauermann, C. Weimann, W. Freude, and C. G. Koos, “Silicon Photonic Optical Coherence Tomography System,” in CLEO: Applications and Technology, (Optical Society of America, 2014), ATu2P. 4.
51. V. Christopoulos, L. Kagemann, G. Wollstein, H. Ishikawa, M. L. Gabriele, M. Wojtkowski, V. Srinivasan, J. G. Fujimoto, J. S. Duker, D. K. Dhaliwal, and J. S. Schuman, “In vivo corneal high-speed, ultra high-resolution optical coherence tomography,” Arch. Ophthalmol. 125(8), 1027–1035 (2007). [CrossRef] [PubMed]
52. M. Wojtkowski, T. Bajraszewski, I. Gorczyńska, P. Targowski, A. Kowalczyk, W. Wasilewski, and C. Radzewicz, “Ophthalmic imaging by spectral optical coherence tomography,” Am. J. Ophthalmol. 138(3), 412–419 (2004). [CrossRef] [PubMed]
53. A. Szkulmowska, M. Szkulmowski, D. Szlag, A. Kowalczyk, and M. Wojtkowski, “Three-dimensional quantitative imaging of retinal and choroidal blood flow velocity using joint Spectral and Time domain Optical Coherence Tomography,” Opt. Express 17(13), 10584–10598 (2009). [CrossRef] [PubMed]
54. D. C. Duffy, J. C. McDonald, O. J. Schueller, and G. M. Whitesides, “Rapid prototyping of microfluidic systems in poly (dimethylsiloxane),” Anal. Chem. 70(23), 4974–4984 (1998). [CrossRef] [PubMed]
55. O. Carvalho, B. Clairac, M. Benderitter, and L. Roy, “Statistical speckle study to characterize scattering media: use of two complementary approaches,” Opt. Express 15(21), 13817–13831 (2007). [CrossRef] [PubMed]
56. Y. Piederrière, J. Cariou, Y. Guern, B. Le Jeune, G. Le Brun, and J. Lortrian, “Scattering through fluids: speckle size measurement and Monte Carlo simulations close to and into the multiple scattering,” Opt. Express 12(1), 176–188 (2004). [CrossRef] [PubMed]
57. A. Dyrnaev, “A method of segmenting leukocytes on images of blood smears,” J. Opt. Technol. 79(11), 708–711 (2012). [CrossRef]