Basic Usage¶
This tutorial covers the essential steps for launching UELer and navigating image data. None of it requires a cell table.
1. Launch the Viewer¶
After configuring your paths, run the launch cells in
script/run_ueler.ipynb. The viewer appears inline in the notebook output.
2. Select an FOV¶
Use the Select Image: dropdown in the left panel to choose a Field of View (FOV). UELer loads image data on demand and caches recently accessed FOVs in memory.
The Cache Size: field controls how many FOVs are held in memory at once (default 100). Lower it if memory is tight; raise it for faster back-and-forth navigation.
3. Select Channels¶
Channels are chosen with the Channels: field. Click it to open a scrollable list of every channel in the dataset, and type to filter that list (case-insensitive). A counter shows how much you are looking at — "12 of 148 shown · 2 selected".
- Select all shown takes the whole filtered group at once, which is the quick way to grab a family
of markers: type
CD, then click it. - Clear drops the entire selection.
- Keyboard: ↑/↓ move, Enter toggles the highlighted channel, Esc closes the list, and Backspace on an empty filter box removes the last channel you added.
Each selected channel then appears as a chip below the field. Drag a chip by its ⋮⋮ grip to
reorder it — a blue bar marks where it will land — or click a chip and press ← / → to move it, or
Delete to remove it. The chip order sets the order of the per-channel control rows, the channel
legend, and the panes of the channel grid view.
The displayed image composites all selected channels using their assigned colors and contrast ranges.
Reordering does not change the image
Channels are composited additively, so the result does not depend on their order and no channel covers another. Reorder for your own convenience — see Display Settings.
4. Load a Marker Set¶
A marker set is a named combination of channels, colors, and contrast ranges. If you have one saved, choose it from the Marker Set: dropdown and click Load Marker Set to restore that configuration instantly.
To save your current configuration, type a name in Set Name: and click Save Marker Set. Use Update Marker Set to overwrite the selected set, or Delete Marker Set (with the Confirm Deletion checkbox) to remove one.
5. Adjust Contrast and Colors¶
Expand the Channels accordion in the left panel. Each selected channel exposes:
- A color dropdown to set its display color.
- A visibility checkbox to show/hide it without deselecting.
- Min and Max contrast sliders to adjust the display range.
These settings are per-session — save them as a marker set to reuse them.
Contrast starts at an automatic range rather than the raw data range, which is what makes a channel look all-black or all-white when the default does not suit it. Display Settings explains where that range comes from and how to correct it.
Enable Show channel legend to display a color key for the visible channels.
6. Overlay Masks and Annotations¶
If masks_folder and/or annotations_folder were provided:
…the left panel grows an accordion section for each:
- In the Masks section, enable a mask via its checkbox and pick a color from its Mask <name> dropdown. Adjust Mask outline px: to change the outline thickness.
- In the Pixel annotations section, enable Show annotation, choose an annotation from the Annotation: dropdown, and adjust Fill alpha: for overlay transparency. Use Edit palette… to customize per-class colors.
If a section is missing, that folder did not yield readable rasters — the sections are created from the data, not toggled on.
For richer per-class mask coloring (fill vs. outline, per-class opacity, continuous colour scales, and saved palettes), use the Mask painter plugin — see the User Interface reference. That one needs a cell table.
Masks-only view
Enable No image (masks only) to hide the channel image and inspect mask/annotation overlays on a blank background.
7. Use the Channel Grid View¶
Enable Channel grid view in the Channels panel to render each visible channel as a separate labelled pane in a synchronized subplot grid — useful for comparing channels side by side without toggling selections.
Next Steps¶
- Explore the full User Interface reference.
- Set the viewer up properly for your dataset — Display Settings.
- Capture Regions of Interest.
- Load a cell table to unlock single-cell analysis.