Bacteria Density Analyzer: User’s guide

0. Load the images and open the plugin’s panel

To use an image, the channels must be separated in different layers of Napari.
The easier way to get that is to have your images split into different files, each file containing one channel.
Another way would be to use napari-imagej (another plugin) that allows Napari to read the memory of Fiji.
If you already have one TIFF file per channel, you can drag’n’drop then into Napari’s main window.
They should appear in the “layers list” (in the lower left corner). The list is a stack, so you only see the uppermost layer.
You can use the “once” button in the “layer controls” panel to adjust the contrast of each layer for better visualization.
You can now go in the top bar menu: Plugins > Density vs. distance to open the plugin’s panel.
The panel should appear on the right side of Napari’s window.

In the “Nuclei channel” dropdown menu, select the layer corresponding to the nuclei channel.

For each channel in which you want to make intensity measurements:
- Use the “Ci” dropdown menu to select the corresponding layer.
- In the text input on the right containing “—” by default, provide the name for this channel that will be used in the results table.
There is always a blank line at the end, it is normal.

Above the “layers list” panel, add a new “Shape layer”.
Add polygons for each region that you would like to process. Before drawing, you can increase the edge thickness for better visibility.
Use the F1, F2, F3 & F4 buttons to change the edge color of these polygons. Each color represents a filament. If several areas share a same color, it means that they are part of the same filament and were cut for some reason (obstacle, crossing, …).
- 1: Red
- 2: Green
- 3: Blue
- 4: Yellow
Add a new “Points layer”. In the “layers control” panel, switch to the “Add points” mode. Before placing the first point, you can increase the size of the points for better visibility.
For each filament (each color), add a point approximately where the filament starts. There should be exactly one point per color.
In the dropdown menus of the ROIs box of the plugin, provide the new shape and point layers.

Using the “Set output folder”, provide the path to an empty directory.
Note: If you already completed an analysis and need to reload the results (for visualization only), you can use this same button and provide the path to the folder that you used for the desired analysis.
Select the binning length using the input below (binning of measures along the skeleton).
Select the list of measurements that you would like to process.

From this point, you can click on each button from top to bottom:
Chunk images: Will export a copy of each selected area in the working directory.
Skeletonize: Will create a mask and a skeleton of the corresponding branch. An outline of the mask will be displayed.
Make medial path: Will process the path over the organ and assemble the fragments. You can check the result by searching for the middle of the stack using the slider under the image.
Measure: Will perform the measures and export the CSV.
Create Plots: Export the plots as PNG with a dashed-line between segments.

Metric	Description
Integrated intensity	Sum of all intensities for the voxels refering to this bin.
Integrated volume	Total volume taken by the voxels refering to this bin.
Local width	Local width measured along the skeleton.

If you need so, you can get an approximation of the discarded volume of organ on your image.
To do that, add a new shape layer and draw a polygon over each area that you didn’t include in any filament.
In the “Discarded polygons” dropdown menu from the plugin’s tab, choose the shape layer containing the polygons that you just drew.
If you click the “Process discarded volume” button, the discarded volume should be displayed right beneath it.
It is an approximation in the way that the area you just indicated cannot be segmented. To estimate the volume, an artificial mask is built. It is as deep (== Z-axis == nb of slices) as the whole stack and has the polygons shape on every slice.