©
Spectradyne LLC, All rights reserved
32
Version 2.5.0 Revision 3, July 2020
2.2
Filter the Set of Detected Particles to Exclude False Positives
Particle detection events manifest as peaks in the raw data. The algorithm used to detect and
characterize these peaks employs a thresholding method: Any excursion of the normalized output
signal crossing above the peak detection threshold is considered to be a possible peak and
measured (see Figure 17). The metrics for each peak found in this way include quality of fit
parameters (i.e., signal-to-noise), as well as physical parameters of the transit event such as transit
time through the sensing constriction and symmetry of the peak shape.
As the peak detection threshold approaches the baseline noise, some false positives (i.e., peaks in
the raw data that do not likely represent real particles) naturally are included in the Stats file data
set (see Figure 17). The goal of this step in the workflow is to eliminate false positives from the
data by excluding particle detection events whose measured peak parameters fall outside the
normal range for true particles. The signal-to-noise, peak symmetry, and transit time parameters
are all available for use as filter parameters.
Figure 17.
The peak detection threshold indicated by the dashed line is close enough to the
baseline noise that two false positives are detected near 200 us and 450 us in the local time window;
a real particle event is clearly detected at 750 us. A red point highlights the peak value, and the
red line shows the fit for each peak. Peak-filtering tools are available in the Viewer for identifying
and removing false positives in the data.