Electrical micro flow cytometry with LSTM and its application in leukocyte differential.

This paper developed an electrical micro flow cytometry to realize leukocyte differentials leveraging a constrictional microchannel and a deep neural network. Firstly, purified granulocytes, lymphocytes or monocytes traveled through the constrictional microchannel with a cross-sectional area marginally larger than individual cells and produced large impedance variations by blocking focused electric field lines. By optimizing key elements (e.g., normalization, learning rate, batch size and neuron number) of the recurrent neural network (RNN), electrical results of purified leukocytes were analyzed to establish a leukocyte differential system with a classification accuracy of 95.2%. Then the leukocyte mixtures were forced to travel through the same constrictional microchannel, producing mixed impedance profiles which were classified into granulocytes, lymphocytes and monocytes based on the aforementioned differential system. As to the classification results, two leukocyte mixtures from the same donor were processed, producing comparable classification results, which were 57% versus 59% of granulocytes, 37% versus 34% of lymphocytes and 6% versus 7% of monocytes. These results validated the established classification system based on the constrictional microchannel and the recurrent neural network, providing a new perspective of differentiating white blood cells by electrical flow cytometry.

Segmentation, feature extraction and classification of leukocytes leveraging neural networks, a comparative study.

The gold standard of leukocyte differentiation is a manual examination of blood smears, which is not only time and labor intensive but also susceptible to human error. As to automatic classification, there is still no comparative study of cell segmentation, feature extraction, and cell classification, where a variety of machine and deep learning models are compared with home-developed approaches. In this study, both traditional machine learning of K-means clustering versus deep learning of U-Net, U-Net + ResNet18, and U-Net + ResNet34 were used for cell segmentation, producing segmentation accuracies of 94.36% versus 99.17% for the dataset of CellaVision and 93.20% versus 98.75% for the dataset of BCCD, confirming that deep learning produces higher performance than traditional machine learning in leukocyte classification. In addition, a series of deep-learning approaches, including AlexNet, VGG16, and ResNet18, was adopted to conduct feature extraction and cell classification of leukocytes, producing classification accuracies of 91.31%, 97.83%, and 100% of CellaVision as well as 81.18%, 91.64% and 97.82% of BCCD, confirming the capability of the increased deepness of neural networks in leukocyte classification. As to the demonstrations, this study further conducted cell-type classification of ALL-IDB2 and PCB-HBC datasets, producing high accuracies of 100% and 98.49% among all literature, validating the deep learning model used in this study.

Leukocyte differential based on an imaging and impedance flow cytometry of microfluidics coupled with deep neural networks.

The differential of leukocytes functions as the first indicator in clinical examinations. However, microscopic examinations suffered from key limitations of low throughputs in classifying leukocytes while commercially available hematology analyzers failed to provide quantitative accuracies in leukocyte differentials. A home-developed imaging and impedance flow cytometry of microfluidics was used to capture fluorescent images and impedance variations of single cells traveling through constrictional microchannels. Convolutional and recurrent neural networks were adopted for data processing and feature extractions, which were then fused by a support vector machine to realize the four-part differential of leukocytes. The classification accuracies of the four-part leukocyte differential were quantified as 95.4% based on fluorescent images plus the convolutional neural network, 90.3% based on impedance variations plus the recurrent neural network, and 99.3% on the basis of fluorescent images, impedance variations, and deep neural networks. Based on single-cell fluorescent imaging and impedance variations coupled with deep neural networks, the four-part leukocyte differential can be realized with almost 100% accuracy.

Advance of microfluidic flow cytometry enabling high-throughput characterization of single-cell electrical and structural properties.

This paper reported a micro flow cytometer capable of high-throughput characterization of single-cell electrical and structural features based on constrictional microchannels and deep neural networks. When single cells traveled through microchannels with constricted cross-sectional areas, they effectively blocked concentrated electric field lines, producing large impedance variations. Meanwhile, the traveling cells were confined within the cross-sectional areas of the constrictional microchannels, enabling the capture of high-quality images without losing focuses. Then single-cell features from impedance profiles and optical images were extracted from customized recurrent and convolution networks (RNN and CNN), which were further fused for cell-type classification based on support vector machines (SVM). As a demonstration, two leukemia cell lines (e.g., HL60 vs. Jurkat) were analyzed, producing high-classification accuracies of 99.3% based on electrical features extracted from Long Short-Term Memory (LSTM) of RNN, 96.7% based on structural features extracted from Resnet18 of CNN and 100.0% based on combined features enabled by SVM. The microfluidic flow cytometry developed in this study may provide a new perspective for the field of single-cell analysis.

Quantitative flow cytometry leveraging droplet-based constriction microchannels with high reliability and high sensitivity.

This study presented a quantitative flow cytometry leveraging droplet-based constriction microchannels with high reliability and high sensitivity. Droplets encapsulating single cells and even distribution of fluorescein labeled antibodies removed from targeted cells deformed through the constriction microchannel where the excited fluorescent signals were sampled and interpreted into numbers of proteins based on volume equivalence in measurement of droplets and calibration of fluorescence. To improve the detection reliability, a comprehensive analysis and comparison of multiple stripping agents such as proteinase K, guanidine hydrochloride, and urea was conducted. To improve the detection sensitivity, light modulation was used to address electrical noises and quartz microchannels were fabricated to address optical noises. As a demonstration, based on this quantitative flow cytometry of droplet microfluidics, (1) mutant p53 expressions of single cells were quantified as 1.95 ± 0.60 × 105 (ncell  = 2918 of A431) and 1.30 ± 0.70 × 105 (ncell  = 3954 of T47D); (2) single-cell expressions of Ras, c-Myc, and ß-tubulin were quantified as 1.90 ± 0.59 × 105 , 4.39 ± 1.44 × 105 , and 2.97 ± 0.81 × 105 (ncell  = 3298 of CAL 27), 1.83 ± 0.58 × 105 , 2.08 ± 0.13 × 106 , and 1.96 ± 0.74 × 105 (ncell  = 5459 of WSU-HN6). As a microfluidic tool capable of quantitatively estimating single-cell protein expressions, this methodology may provide a new quantitative perspective for the field of flow cytometry.

Toward five-part differential of leukocytes based on electrical impedances of single cells and neural network.

The five-part differential of leukocytes plays key roles in the diagnosis of a variety of diseases and is realized by optical examinations of single cells, which is prone to various artifacts due to chemical treatments. The classification of leukocytes based on electrical impedances without cell treatments has not been demonstrated because of limitations in approaches of impedance acquisition and data processing. In this study, based on treatment-free single-cell impedance profiles collected from impedance flow cytometry leveraging constriction microchannels, two types of neural pattern recognition were conducted for comparisons with the purpose of realizing the five-part differential of leukocytes. In the first approach, 30 features from impedance profiles were defined manually and extracted automatically, and then a feedforward neural network was conducted, producing a classification accuracy of 84.9% in the five-part leukocyte differential. In the second approach, a customized recurrent neural network was developed to process impedance profiles directly and based on deep learning, a classification accuracy of 97.5% in the five-part leukocyte differential was reported. These results validated the feasibility of the five-part leukocyte differential based on label-free impedance profiles of single cells and thus provide a new perspective of differentiating white blood cells based on impedance flow cytometry.

Advance of microfluidic constriction channel system of measuring single-cell cortical tension/specific capacitance of membrane and conductivity of cytoplasm.

This paper reported a microfluidic platform which realized the characterization of inherent single-cell biomechanical and bioelectrical parameters simultaneously. Individual cells traveled through a constriction channel with deformation images and impedance variations captured and processed into cortical tension Tc , specific membrane capacitance Csm , and cytoplasmic conductivity σcy based on an equivalent biophysical model. These properties of thousands of individual cells of K562, Jurkat, HL-60, HL-60 treated with paraformaldehyde (PA)/cytochalasin D (CD)/concanavalin A (ConA), granulocytes of Donor 1, Donor 2, and Donor 3 were quantified for the first time. Leveraging Tc , Csm , and σcy , (1) high accuracies of classifying wild-type and processed HL-60 cells (e.g., 93.5% of PA treated vs. CD treated HL-60 cells) were realized, revealing the effectiveness of using these three biophysical parameters in cell-type classification; (2) low accuracies of classifying normal granulocytes from three donors (e.g., 56.4% of Donor 1 vs. 2), indicating comparable parameters for normal granulocytes. In conclusion, this platform can characterize single-cell Tc , Csm , and σcy concurrently and quantify multiple parameters in single-cell analysis.

Inherent single-cell bioelectrical parameters of thousands of neutrophils, eosinophils and basophils derived from impedance flow cytometry.

Single-cell bioelectrical properties are commonly used for blood cell phenotyping in a label-free manner. However, previously reported inherent single-cell bioelectrical parameters (e.g., diameter Dc , specific membrane capacitance Csm and cytoplasmic conductivity σcy ) of neutrophils, eosinophils and basophils were obtained from only tens of individual cells with limited statistical significance. In this study, granulocytes were separated into neutrophils, eosinophils and basophils based on fluorescent flow cytometry, which were further aspirated through a constriction-microchannel impedance flow cytometry for electrical property characterization. Based on this microfluidic impedance flow cytometry, single-cell values of Dc , Csm and σcy were measured as 10.25 ± 0.66 µm, 2.17 ± 0.30 µF/cm2 , and 0.37 ± 0.05 S/m for neutrophils (ncell  = 9442); 9.73 ± 0.51 µm, 2.07 ± 0.19 µF/cm2 , and 0.30 ± 0.04 S/m for eosinophils (ncell  = 2982); 9.75 ± 0.49 µm, 2.06 ± 0.17 µF/cm2 , and 0.31 ± 0.04 S/m for basophils (ncell  = 5377). Based on these inherent single-cell bioelectrical parameters, neural pattern recognition was conducted, producing classification rates of 80.8% (neutrophil vs. eosinophil), 77.7% (neutrophil vs. basophil) and 59.3% (neutrophil vs. basophil). These results indicate that as inherent single-cell bioelectrical parameters, Dc , Csm and σcy can be used to classify neutrophils from eosinophils or basophils to some extent while they cannot be used to effectively distinguish eosinophils from basophils.

Inherent bioelectrical parameters of hundreds of thousands of single leukocytes based on impedance flow cytometry.

As label-free biomarkers, bioelectrical properties of single cells have been widely used in hematology analyzers for 3-part differential of leukocytes, in which, however, instrument dependent bioelectrical parameters (e.g., DC/AC impedance values) rather than inherent bioelectrical parameters (e.g., diameter Dc , specific membrane capacitance Csm and cytoplasmic conductivity σcy ) were used, leading to poor comparisons among different instruments. In order to address this issue, this study collected inherent bioelectrical parameters from hundreds of thousands of white blood cells based on a home-developed impedance flow cytometry with corresponding 3-part differential of leukocytes realized. More specifically, leukocytes were separated into three major subtypes of granulocytes, monocytes and lymphocytes based on density gradient centrifugation. Then these separated cells were aspirated through a constriction-microchannel based impedance flow cytometry where inherent bioelectrical parameters of Dc , Csm and σcy were quantified as 9.8 ± 0.7 µm, 2.06 ± 0.26 µF/cm2 , and 0.34 ± 0.05 S/m for granulocytes (ncell  = 134,829); 10.4 ± 1.0 µm, 2.45 ± 0.48 µF/cm2 , and 0.42 ± 0.08 S/m for monocytes (ncell  = 40,226); 8.0 ± 0.5 µm, 2.23 ± 0.34 µF/cm2 , and 0.35 ± 0.08 S/m for lymphocytes (ncell  = 129,193). Based on these inherent bioelectrical parameters, neural pattern recognition was conducted, producing a high "classification accuracy" of 93.5% in classifying these three subtypes of leukocytes. These results indicate that as inherent bioelectrical parameters, Dc , Csm , and σcy can be used to electrically phenotype white blood cells in a label-free manner.

MEMS-Based Electrochemical Seismometer Relying on a CAC Integrated Three-Electrode Structure.

This study developed a MEMS-based electrochemical seismometer relying on a cathode-anode-cathode (CAC) integrated three-electrode structure where two cathodes were positioned on two surfaces of a silicon wafer, while one anode was positioned on the sidewalls of the through holes of the silicon wafer. Device design and numerical simulations were conducted to model the functionality of the three-electrode structure in detecting vibration signals with the key geometrical parameters optimized. The CAC integrated three-electrode structure was then manufactured by microfabrication, which demonstrated a simplified fabrication process in comparison with conventional four-electrode structures. Device characterization shows that the sensitivity of the CAC microseismometer was an order of magnitude higher than that of the CME6011 (a commercially available four-electrode electrochemical seismometer), while the noise level was comparable. Furthermore, in response to random vibrations, a high correlation coefficient between the CAC and the CME6011 (0.985) was located, validating the performance of the developed seismometer. Thus, the developed electrochemical microseismometer based on an integrated three-electrode structure may provide a new perspective in seismic observations and resource explorations.

Microfluidic Cytometry for High-Throughput Characterization of Single Cell Cytoplasmic Viscosity Using Crossing Constriction Channels.

This article presents an approach of microfluidic flow cytometry capable of continuously characterizing cytoplasmic viscosities of single cells. The microfluidic system consists of a major constriction channel and a side constriction channel perpendicularly crossing each other. Cells are forced to rapidly travel through the major channel and are partially aspirated into the side channel when passing the channel junction. Numerical simulations were conducted to model the time dependence of the aspiration length into the side channel, which enables the measurement of cytoplasmic viscosity by fitting the model results to experimental data. As a demonstration for high-throughput measurement, the cytoplasmic viscosities of HL-60 cells that were native or treated by N-Formylmethionine-leucyl-phenylalanine (fMLP) were quantified with sample sizes as large as thousands of cells. Both the average and median cytoplasmic viscosities of native HL-60 cells were found to be about 10% smaller than those of fMLP-treated HL-60 cells, consistent with previous observations that fMLP treatment can increase the rigidity of white blood cells. Furthermore, the microfluidic system was used to process granulocytes from three donors (sample size >1,000 cells for each donor). The results revealed that the cytoplasmic viscosity of granulocytes from one donor was significantly higher than the other two, which may result from the fact that this donor just recovered from an inflammation. In summary, the developed microfluidic system can collect cytoplasmic viscosities from thousands of cells and may function as an enabling tool in the field of single-cell analysis. © 2019 International Society for Advancement of Cytometry.

Crossing constriction channel-based microfluidic cytometry capable of electrically phenotyping large populations of single cells.

This paper presents a crossing constriction channel-based microfluidic system for high-throughput characterization of specific membrane capacitance (Csm) and cytoplasm conductivity (σcy) of single cells. In operations, cells in suspension were forced through the major constriction channel and instead of invading the side constriction channel, they effectively sealed the side constriction channel, which led to variations in impedance data. Based on an equivalent circuit model, these raw impedance data were translated into Csm and σcy. As a demonstration, the developed microfluidic system quantified Csm (3.01 ± 0.92 µF cm-2) and σcy (0.36 ± 0.08 S m-1) of 100 000 A549 cells, which could generate reliable results by properly controlling cell positions during their traveling in the crossing constriction channels. Furthermore, the developed microfluidic impedance cytometry was used to distinguish paired low- and high-metastatic carcinoma cell types of SACC-83 (ncell = ∼100 000) and SACC-LM cells (ncell = ∼100 000), distinguishing significant differences in both Csm (3.16 ± 0.90 vs. 2.79 ± 0.67 µF cm-2) and σcy (0.36 ± 0.06 vs.0.41 ± 0.08 S m-1). As high-throughput microfluidic impedance cytometry, this technique may add a new marker-free dimension to flow cytometry in single-cell analysis.

Microelectromechanical System-Based Electrochemical Seismometers with Two Pairs of Electrodes Integrated on One Chip.

This paper presents microelectromechanical system (MEMS)-based electrochemical seismometers with two pairs of electrodes integrated on one chip. Both theoretical analysis and numerical simulations were conducted to reveal the working principle of the proposed electrochemical seismometers, finding that flow holes distributed on cathodes rather than anodes can produce significantly higher sensitivities. The proposed electrochemical seismometers were fabricated based on conventional micromachined processes with high fabrication repeatability. Sensitivity measurements of the proposed seismometers and their commercial counterpart of CME6011 were conducted, indicating the sensitivities of the proposed seismometer with flow holes on cathodes were two orders higher than the counterpart with flow holes on anodes and one order higher than CME6011 at dominant frequencies. Measurements of random ground motions based on the proposed seismometer with flow holes on cathodes and CME6011 were conducted, producing comparable noise levels without observable ground motions and high correlations in response to random events of ground motions. These results validated the functionality of the proposed electrochemical seismometers, which may contribute to seismic monitoring in the near future.

A Temperature-Insensitive Resonant Pressure Micro Sensor Based on Silicon-on-Glass Vacuum Packaging.

This paper presents a temperature-insensitive resonant pressure sensor, which is mainly composed of a silicon-on-insulator (SOI) wafer for pressure measurements and a silicon-on-glass (SOG) cap for vacuum packaging. The variations of pressure under measurement bend the pressure sensitive diaphragm and regulate the intrinsic frequencies of the resonators in the device layer. While, variations of temperature cannot significantly change the intrinsic frequencies of the resonators, due to the SOG cap to offset generated thermal stress. Numerical simulations, based on finite element analysis, were conducted to calculate the residual thermal stress and optimize the sensing structures. Experimental results show that the Q-factors of the resonators are higher than 16,000, with a differential pressure sensitivity of 11.89 Hz/kPa, a nonlinearity of 0.01% F.S and a low fitting error of 0.01% F.S with the pressure varying from 100 kPa to 1000 kPa. In particular, a temperature sensitivity of ~1 Hz/°C was obtained in the range of -45 °C to 65 °C, which is one order of magnitude lower than the previously reported counterparts.

A Resonant Pressure Microsensor with the Measurement Range of 1 MPa Based on Sensitivities Balanced Dual Resonators.

This paper presents a resonant pressure microsensor with the measurement range of 1 MPa suitable for the soaring demands of industrial gas pressure calibration equipment. The proposed microsensor consists of an SOI layer as a sensing element and a glass cap for vacuum packaging. The sensing elements include a pressure-sensitive diaphragm and two resonators embedded in the diaphragm by anchor structures. The resonators are excited by a convenient Lorentz force and detected by electromagnetic induction, which can maintain high signal outputs. In operation, the pressure under measurement bends the pressure-sensitive diaphragm of the microsensor, producing frequency shifts of the two underlining resonators. The microsensor structures were designed and optimized using finite element analyses and a 4" SOI wafer was employed in fabrications, which requires only one photolithographic step. Experimental results indicate that the Q-factors of the resonators are higher than 25,000 with a differential temperature sensitivity of 0.22 Hz/°C, pressure sensitivities of 6.6 Hz/kPa, and -6.5 Hz/kPa, which match the simulation results of differential temperature sensitivity of 0.2 Hz/°C and pressure sensitivities of ±6.5 Hz/kPa. In addition, characterizations based on a closed-loop manner indicate that the presented sensor demonstrates low fitting errors within 0.01% FS, high accuracy of 0.01% FS in the pressure range of 20 kPa to 1 MPa and temperature range of -55 to 85 °C, and the long-term stability within 0.01% FS in a 156-day period under the room temperature.

Mechanical property characterization of hundreds of single nuclei based on microfluidic constriction channel.

As label-free biomarkers, the mechanical properties of nuclei are widely treated as promising biomechanical markers for cell type classification and cellular status evaluation. However, previously reported mechanical parameters were derived from only around 10 nuclei, lacking statistical significances due to low sample numbers. To address this issue, nuclei were first isolated from SW620 and A549 cells, respectively, using a chemical treatment method. This was followed by aspirating them through two types of microfluidic constriction channels for mechanical property characterization. In this study, hundreds of nuclei were characterized, producing passage times of 0.5 ± 1.2 s for SW620 nuclei in type I constriction channel (n = 153), 0.045 ± 0.047 s for SW620 nuclei in type II constriction channel (n = 215) and 0.50 ± 0.86 s for A549 nuclei in type II constriction channel. In addition, neural network based pattern recognition was used to classify the nuclei isolated from SW620 and A549 cells, producing successful classification rates of 87.2% for diameters of nuclei, 85.5% for passage times of nuclei and 89.3% for both passage times and diameters of nuclei. These results indicate that the characterization of the mechanical properties of nuclei may contribute to the classification of different tumor cells.

A Resonant Pressure Microsensor Based on Double-Ended Tuning Fork and Electrostatic Excitation/Piezoresistive Detection.

This paper presents a resonant pressure microsensor relying on electrostatic excitation and piezoresistive detection where two double-ended tuning forks were used as resonators, enabling differential outputs. Pressure under measurement caused the deformation of the pressure sensitive membrane, leading to stress buildup of the resonator under electrostatic excitation with a corresponding shift of the resonant frequency detected piezoresistively. The proposed microsensor was fabricated by simplified SOI-MEMS technologies and characterized by both open-loop and closed-loop circuits, producing a quality factor higher than 10,000, a sensitivity of 79.44 Hz/kPa and an accuracy rate of over 0.01% F.S. In comparison to the previously reported resonant piezoresistive sensors, the proposed device used single-crystal silicon as piezoresistors, which was featured with low DC biased voltages, simple sensing structures and fabrication steps. In addition, the two double-ended tuning forks were used as resonators, producing high quality factors and differential outputs, which further improved the sensor performances.

A Monolithic Electrochemical Micro Seismic Sensor Capable of Monitoring Three-Dimensional Vibrations.

A monolithic electrochemical micro seismic sensor capable of monitoring three-axial vibrations was proposed in this paper. The proposed micro sensor mainly consisted of four sensing units interconnected within flow channels and by interpreting the voltage outputs of the sensing units, vibrations with arbitrary directions can be quantified. The proposed seismic sensors are fabricated based on MEMS technologies and characterized, which produced sensitivities along x, y, and z axes as 2473.2 ± 184.5 V/(m/s), 2261.7 ± 119.6 V/(m/s), and 3480.7 ± 417.2 V/(m/s) at 30 Hz. In addition, the vibrations in x-y, x-z, and y-z planes were applied to the developed seismic sensors, leading to comparable monitoring results after decoupling calculations with the input velocities. Furthermore, the results have shown its feasibilities for seismic data recording.

A Flexible Sensing Unit Manufacturing Method of Electrochemical Seismic Sensor.

This paper presents an electrochemical seismic sensor in which paraylene was used as a substrate and insulating layer of micro-fabricated electrodes, enabling the detection of seismic signals with enhanced sensitivities in comparison to silicon-based counterparts. Based on microfabrication, paralene-based electrochemical seismic sensors were fabricated in which the thickness of the insulating spacer was 6.7 µm. Compared to silicon-based counterparts with ~100 µm insulating layers, the parylene-based devices produced higher sensitivities of 490.3 ± 6.1 V/(m/s) vs. 192.2 ± 1.9 V/(m/s) at 0.1 Hz, 4764.4 ± 18 V/(m/s) vs. 318.9 ± 6.5 V/(m/s) at 1 Hz, and 4128.1 ± 38.3 V/(m/s) vs. 254.5 ± 4.2 V/(m/s) at 10 Hz. In addition, the outputs of the parylene vs. silicon devices in response to two transit inputs were compared, producing peak responses of 2.97 V vs. 0.22 V and 2.41 V vs. 0.19 V, respectively. Furthermore, the self-noises of parylene vs. silicon-based devices were compared as follows: -82.3 ± 3.9 dB vs. -90.4 ± 9.4 dB at 0.1 Hz, -75.7 ± 7.3 dB vs. -98.2 ± 9.9 dB at 1 Hz, and -62.4 ± 7.7 dB vs. -91.1 ± 8.1 dB at 10 Hz. The developed parylene-based electrochemical seismic sensors may function as an enabling technique for further detection of seismic motions in various applications.

A microfabricated 96-well wound-healing assay.

This article presents a microfabricated 96-well wound-healing assay enabling high-throughput measurement of cellular migration capabilities. Within each well, the middle area is the wound region, made of microfabricated gold surface with self-assembled PEG repellent for cell seeding. After the formation of a cellular confluent monolayer around the wound region, collagen solution was applied to form three-dimensional matrix to cover the PEG surface, initiating the wound-healing process. By interpreting the numbers of migrated cells into the wound regions as a function of specific stimuli with different concentrations, EC50 (half-maximal effective concentration) was obtained. Using H1299 as a model, values of EC50 were quantified as 8% and 160 ng/ml for fetal bovine serum and CXCL12, respectively. In addition, the values of EC50 were demonstrated not to be affected by variations in compositions of extracellular matrix and geometries of wounds, which can thus be regarded as an intrinsic marker. Furthermore, the migration capabilities of a second cell type (HeLa) were characterized by the developed wound-healing assay, producing EC50 of 2% when fetal bovine serum was used as the stimuli. These results validated the proposed high-throughput wound-healing assay, which may function as an enabling tool in studying cellular capabilities of migration and invasion. © 2017 International Society for Advancement of Cytometry.