TECHNICAL NOTE

Using File Processing Mode for FlowCam Method Development

INTRODUCTION

During flow imaging microscopy, sample volume flows continuously

through a flow cell between an LED light source and a digital camera,
which captures images of the flow cell’s field of view at regular
intervals. These images are referred to as raw camera images in
VisualSpreadsheet and in this document. Particle images are then
segmented from these raw camera images according to capture
settings defined in the Context dialog. Because each FlowCam
application deals with different particle types that exhibit unique
optical properties, users need to perform method development
For a detailed practical walk-through of using File Processing Mode
when getting started with FlowCam and when analyzing new
to optimize segmentation, please refer to our Capture Settings
particle types to ensure that they obtain accurate size, morphology,
for Biopharmaceutical Applications—Part 2: Optimizing Settings
and concentration measurements for their specific particle types.
technical note.
VisualSpreadsheet software is a powerful tool that allows users to
customize and adjust the capture settings used to detect and image
particles during method development. File Processing Mode, a
SAVING RAW IMAGE DATA
feature available on all FlowCam models, is a convenient software
tool that optimizes capture settings for specific applications and The first step in using File Processing Mode to optimize capture
sample types. It allows users to re-process saved raw camera images settings is to analyze relevant samples on a properly configured
from a single sample run at different capture settings, eliminating FlowCam instrument. While doing so, it’s essential to save raw camera
the need to analyze multiple sample aliquots to fine-tune these images from the analysis. These raw camera images can then be re-
parameters. This technical note details the process of saving raw processed at different capture settings using File Processing Mode,
camera images and utilizing File Processing Mode to adjust the allowing you to assess the impact of these settings on measurements
capture settings, ensuring efficient and accurate concentration, without the variability of different samples affecting the analysis.
sizing, and morphology measurements with FlowCam. This process
consists of 4 steps:
To collect image data appropriate for capture setting adjustments:
1. Saving Raw Image Data: These are the raw camera images that
will be re-processed in the event optimization is necessary. 1. Install the objective for which the capture settings are to be optimized.
Then, install the corresponding syringe and flow cell for that objective.
2. Evaluating Particle Images: This is a qualitative assessment of
segmentation quality to assess if optimization is necessary and 2. Autofocus the instrument using an appropriate polystyrene
to determine which capture settings to adjust. bead solution.

3. Adjusting Capture Settings: Users iteratively adjust specific 3. After successfully autofocusing, flush out any remaining bead
capture settings prior to performing File Processing mode. solution, followed by at least 1 mL of particle-free water. Flush
all fluid out of the flow cell before proceeding.
4. Re-Analyzing Raw Images with File Processing Mode: The raw
images are re-segmented with the adjusted capture settings 4. From the main VisualSpreadsheet window, select Context and
from Step 3 and can then be qualitatively reevaluated. navigate to the Capture tab. For FlowCam 8000 Series instruments, we
recommend starting with the settings shown in Table 1 as a baseline.

| Yokogawa Fluid Imaging Technologies, Inc. | www.flowcam.com | +1-207-289-3200 1

 Plankton Biopharma & 6. Verify that the settings in the Sample Settings, Capture Filter,
Applications Materials Applications and Stop tabs are set appropriately for your protocol.
(color camera) (grayscale camera)
7. Use the pipettor to load an appropriate volume of a representative
Objective Lens 20X 10X 4X 20X 10X 4X sample into the sample port.

Distance to 8. Analyze the sample in AutoImage mode.

8 8 10 0 0 0
Nearest Neighbor
9. Flush the system with at least 1 mL of particle-free water to clean
Close Holes 1 1 1 3 3 3
the flow cell.

Dark Pixels 17 17 20 12 12 12

Particles defined by
Light Pixels 17 17 20
dark pixels only EVALUATING PARTICLE IMAGES FROM INITIAL DATA
Table 1. Recommended starting capture settings for File Processing Mode by
CAPTURE
Application/Camera Type
Once you’ve completed a data run with raw image capture enabled,
select Open View to evaluate the particle images collected during
5. While still in the Capture tab, ensure that the Save Raw Camera the measurement. From the view window, toggle on the particle
Images checkbox is checked and Max Raw Camera Images is edge trace overlay (Show > Particle Edge Trace) to see where pixels
set to a high value (e.g., 5000), as shown in Figure 1. Click File > were detected in each image and assess the segmentation accuracy.
Save As and save this context as ObjX_General Settings, adding This step is significant as it helps you determine if further setting
the objective magnification in place of the Obj placeholder. optimization is required. If the edge trace qualitatively agrees with
the outer particle edge for most images, as shown in Figure 2, then
your initial data capture was successful.

Figure 2. The red line indicates the particle edge trace and aligns closely with the
outer particle edge. In this case, no setting optimization is necessary.

However, if you notice image artifacts like those shown in Table 2

(next page), then File Processing Mode should be used to optimize
your capture settings, as these artifacts can impact size and
concentration measurements. Table 2 provides examples of each
artifact type, their impacts on measurements, and recommended
capture setting adjustments to reduce the incidence of such
Figure 1. Capture settings tab showing generalized settings and Save Raw Camera
Images enabled artifacts.

2
 Recommended
Example Images Impact on Measurements Capture Setting
Adjustment

a) Blank • Concentration will be artificially

increased due to the inclusion of blank ↑ Increase Dark
background
images and/or Light Pixel
images • Errors in summary statistics like Thresholds
average particle size

• Particles will be oversized because

the background is included in the edge
↑ Increase Light
b) Haloing trace
Pixel Thresholds
• Errors in other particle property
measurements

1. ↓ Decrease
Dark and/or Light
c) Holes are • Particles will be undersized because
Pixel Thresholds
holes are not included in calculations
present in the • Errors in other particle property
depending on which
segmentation pixels are omitted
measurements
2. ↑ Increase Close
Holes

1. ↓ Decrease Dark
• Particles will be undersized and
and/or Light Pixel
overcounted as a result of fragmenting
Thresholds
a single particle into multiple smaller
d) Fragmentation pieces
2. ↑ Increase Close
Holes
• It will be challenging to classify
3. ↑ Increase Distance
particles appropriately
to Nearest Neighbor

• Particles will be oversized and

undercounted as a result of combining
e) Multiple 1. ↓ Decrease Close
two particles in one image
Holes
particles • It will be difficult to classify particles
2. ↓ Decrease Distance
counted as one appropriately when two different
to Nearest Neighbor
organisms or particle types are
combined

• ↓ Decrease
f) Missing Dark and/or Light
Pixel Thresholds
particles from • Specific particle types will be
depending on which
raw camera undercounted
particles are ignored
images • Check Capture Filter
settings

Table 2. Examples of sub-optimal capture settings, their impact on measurements, and suggested capture setting adjustments. In a), blank background images are being
detected. Compare the edge trace in b) to Figure 2 and note the extensive haloing occurring here. Compare the edge trace in c) to Figure 2, and note the holes present here.
In d), a single Fragilaria diatom has been segmented into multiple image fragments, and in e), multiple particles are being counted as one particle. In f), the small particle in
the lower right was not segmented from the raw camera image and thus not counted.

3
ADJUSTING THE CAPTURE SETTINGS 2. Click the File Processing Mode button from the operations bar in
the main window of VisualSpreadsheet (FlowCam version) or click
To optimize capture settings using File Processing Mode most
Analyze > File Processing Mode (Satellite version) (Figure 4).
efficiently, we recommend making modifications in the order and
increments shown in Table 3, after starting with the context settings
previously shown in Table 1.

# to raise or low-
Adjustment
Property er the value in
order
each iteration
1 Dark Pixel Threshold 2

2 Light Pixel Threshold 5 Figure 4. Enabling File Processing Mode on FlowCam (top) and on a satellite
computer (Bottom)
3 Close Holes (Iterations) 1
Distance to Nearest Neighbor
4 1
(µm) 3. A File Explorer Window will appear. Navigate to the location
Table 3. Order and increment in which the capture settings should be adjusted of your saved raw image data—the default location is
during File Processing Mode DocumentsVisualSpreadsheet6 data. Select the first raw image
file, which will be named rawfile_000000, and click Open (Figure 5).

RE-ANALYZING RAW IMAGES WITH FILE PROCESSING

MODE
Once you’ve assessed your images and decided that capture setting
optimization is necessary, you are ready to use File Processing Mode.
The steps below walk through one iteration of the process:

1. Change one of the capture settings shown in Table 3. This

example follows the suggested order and adjusts Dark Pixels,
raising it from the original recommended setting by 2 (Figure 3).

Figure 5. Selecting rawfile_000000 to initiate File Processing Mode data acquisition

4. You will be prompted to provide stop conditions and a run name.

When naming the run, keep the original run name and then add
which setting was changed so you can keep track of each run. Since
the dark pixels setting was raised to 14 in this example, adding DP
14 to the end of the original name would be appropriate.

5. Like a standard AutoImage mode measurement, the capture

window and the particle collage window will be displayed. Boxes
will appear around any particle detected by the segmentation
algorithm in the raw camera images.

6. At the completion of the run, click Open View to evaluate your

particle images and edge traces as covered on page 2. It may be
helpful to open the original run and the file processing run images
Figure 3. Raising Dark Pixels from 12 to 14 prior to utilizing File Processing Mode side by side to compare them directly (Figure 6).

4
 7. Once the dark pixel threshold has been optimized, continue
to move through the settings in Table 3 until the particle edge
trace closely aligns with the true particle edge and you don’t see
the artifacts described in Table 2. Once your settings have been
optimized, make sure to save your new context file so it can be
used for future data acquisition.

8. Once method development is complete, we recommend

unchecking the Save Raw Camera images option in Capture
settings. Now that your Capture settings are optimized, there is
no need to continue saving these large image files as each set of
raw camera images consumes multiple gigabytes worth of hard
drive space that could otherwise be used for data acquisition.

SUMMARY
FlowCam instruments are flexible platforms that can accurately
analyze a variety of samples and particle types in wide-ranging
applications. File Processing Mode is an easy-to-use technique for
method development that allows users to capitalize on this flexibility
Figure 6. Comparison of images captured from original run (top) to those where
and efficiently optimize capture settings for specific sample and
Dark Pixel Threshold was increased to 14 (bottom). In this case, the particle edge
trace is more accurate for the original images than for the re-processed images particle types without the need to re-run samples. Imaging particles
where portions of the exterior and interior regions of particles were omitted (yellow with capture settings that have been optimized through file
arrows). Note the decrease in Diameter ABD shown below the images because of
these missing regions. processing mode ensures that users are collecting accurate sizing,
morphological, and concentration data.

Have questions or need support?

Contact us at [email protected]

Get in touch
1 Tell us about your needs and how we can
support you

We’ll assist you

2 We will get back to you to answer your questions
or schedule support or services

© Yokogawa Fluid Imaging Technologies – All Rights Reserved 5