Fig 1.
Diagrams of convolution of laser profile with an example particle and the resulting voltage versus time.
a) Shows an example particle and its resulting voltage versus time. For purposes of illustration, this can be considered a reference signal. Measurements of pulse height (H), width (W) and area (green under the curve) are highlighted. b) Changes in the velocity primarily affects the width of the signal. These effects are characterized by the matrix in Eq. (1). c) Difference in size affects both the height and width. These effects are characterized by the matrix
in Eq. (1). d) Difference in the dye concentration primarily affects the height of the pulse, which is characterized by the factor of
in Eq. (1). The main idea behind our signals analysis is to determine the set of physical parameters (particle size and speed) that can be used to transform the measured signal into the reference by minimizing the difference
; see Eq. (1). Note that differences in size can only be detected if the particle radius is on the same order of magnitude as the width of the laser profile.
Fig 2.
STA analysis based on reference signal choice. a) Overlay of time-aligned, raw time-series of bead measurements. Colored curves indicate various reference signals that were tested. b) The relative radius distributions of the same bead population are overlaid given reference signal choice. c) Box and whisker plots of the radii for each reference signal after normalizing the mean of each distribution to 1. Throughout this figure the median refers to the pointwise median.
Fig 3.
STA analysis of particles having different velocities.
a) Plot of the velocities of bright, 15 µm diameter particles based on time of flight between two measurement regions over time. Each dot represents a single particle, with the colors representing the clustering of the data. Beads that were collected while the time-averaged velocity was 0.390 m/s (fast) are shown in black, beads collected during transition to slower velocity are shown in maroon, while beads that were collected while the time-averaged velocity was 0.204 m/s (slow) are shown in red. Grey dots indicate that the beads were not used in the analysis. b) Overlay of raw signal pulses from beads collected at different total flow rates. c) Relative time-averaged velocities versus relative extracted velocities. The gray dashed grey line shows a liner regression with an equation of y = 0.99x + 0.014 and correlation coefficient, R2 = 0.999. Histograms of extracted velocities and time-averaged velocities are shown above and to the right of the main figure, respectively. d) Overlay of histograms of the relative radii extracted from STA from different conditions. Bin width is to scale on axis.
Fig 4.
Beads with different sizes but similar dye concentrations were measured and analyzed with the STA.
a) Overlays of fluorescent pulses of the selected subpopulations, black indicates the 10.5 µm diameter bead and blue indicates the 6.1 µm diameter bead. b) Histogram of the relative particle radius of the measured beads.
Fig 5.
Comparison of beads with different fluorescence intensity values.
a) Fluorescence pulses of the beads for the two different 10.5 µm bead intensities (MEFL of 381106 (black) and 128924 (brown)). b) Overlays of the pulses after applying a correction factor to the population with the smaller MEFL. c) Histogram of the two bead populations relative particle radius showing a distribution of similar width.
Fig 6.
Hoechst 33342 labeled Jurkat cells measured in a serial microcytometer.
a) Histogram of the fluorescent area used to gate cells. b) Raw fluorescence pulses for cells gated from the histogram. c) Histogram of nuclei relative radii extracted from STA. Cells likely in G1/G0 (1x DNA) labeled dark gray and cells likely in G2/M (2x DNA) labeled in green.