Spots¶
The Spots tab provides spot detection and colocalization visualization for analyzing fluorescent spots in your images.
Interface overview¶
Spot detection section¶
Controls:
- Image layer (dropdown): Select the image layer containing spots to detect.
- Detector (dropdown): Choose a spot detection algorithm.
- Settings (dynamic panel): Detector-specific parameters.
- Minimum diameter (spin box): Minimum spot diameter in pixels (
0= no minimum filter). - Maximum diameter (spin box): Maximum spot diameter in pixels (
0= no maximum filter). - Run (button): Execute spot detection on the selected image.
Diameter filtering behavior: After detection, connected components are filtered against diameter-derived thresholds:
- For 2D labels, thresholds are converted to effective area:
A = pi * (d / 2)^2. - For 3D labels, thresholds are converted to effective volume:
V = (4/3) * pi * (d / 2)^3.
Components with pixel/voxel counts outside the computed range are removed.
Colocalization section¶
Controls:
- Labels A (dropdown): First labels layer.
- Labels B (dropdown): Second labels layer.
- Visualize (button): Compute intersection and visualize overlaps.
Output:
Creates a Points layer named {labels_A}_{labels_B}_colocalization with yellow ring markers at overlap locations.
Available detectors¶
| Detector | Algorithm | Description |
|---|---|---|
rmp |
Rotational Morphological Processing (RMP) | Spot extraction with rotating images and thin structuring elements. Compatible with 2D and 3D images. |
ufish |
U-FISH | Spot extraction with a compact deep learning model for 2D and 3D images. |
Output layers¶
- Spot detection outputs
<image layer>_<detector>_spot_labels.
Detector settings¶
rmp¶
Source: src/senoquant/tabs/spots/models/rmp/details.json and src/senoquant/tabs/spots/models/rmp/model.py.
Method reference: Rotational Morphological Processing for spot detection.
| Setting | Type | Default | Range | Description |
|---|---|---|---|---|
| Spot diameter (px) | int | 10 | 3 - 9999 | Expected spot diameter used by the extraction structuring element. |
| Auto threshold | bool | true | n/a | Uses Otsu thresholding on the normalized response. |
| Manual threshold | float | 0.50 | 0.0 - 1.0 | Fixed threshold when Auto threshold is off. Disabled when auto-thresholding is enabled. |
RMP simplifications:
- Angle spacing is fixed internally to 5 degrees.
- Wavelet denoising is always enabled (both before and after top-hat extraction).
ufish¶
Source: src/senoquant/tabs/spots/models/ufish/details.json and src/senoquant/tabs/spots/models/ufish/model.py.
| Setting | Type | Default | Range | Description |
|---|---|---|---|---|
| Spot size | float | 1.0 | 0.25 - 4.0 | Spot-scale control. 1.0 is default, >1 biases detection toward larger spots, <1 toward smaller spots. |
| Threshold | float | 0.5 | 0.0 - 1.0 | Foreground threshold on the enhanced response (lower = more spots). |
UFISH simplification: - Wavelet denoising is always enabled before enhancement/segmentation.