Volume To Mask

The Volume To Mask tool converts volume image data into mask objects. This is useful for creating masks from pre-segmented data, threshold-based volumes, or for direct data type conversion.

Accessing the Tool

1. Navigate to the Image tab in the ribbon.
2. Click Volume To Mask in the Convert section.

Conversion Modes

The tool offers two conversion modes:

Direct Copy Mode

Creates a mask by directly copying volume voxel values.

Behavior	Description
8-bit data	Copied directly to the mask
Non-8-bit data	Rescaled to 0–255 range before copying

Use case: Quick conversion when the volume already contains label-like data.

Segmentation Mode

Creates masks from pre-segmented volumes by extracting labels within a threshold range.

Parameter	Description
Lower threshold	Minimum intensity value to include
Upper threshold	Maximum intensity value to include
Multi-label mask	Create one mask with multiple labels or separate masks

Use case: Extracting specific labels from AI-segmented or atlas data.

---

Parameters

Target Objects

Select which volumes to convert:

• Active volume object
• Selected volume objects
• All volume objects

Mode Selection

Mode	Description
Direct copy	Copy volume values directly to mask
Segmentation	Extract labels using threshold range

Segmentation Options

When using Segmentation mode:

Parameter	Description
Lower threshold	Include voxels with values ≥ this threshold
Upper threshold	Include voxels with values ≤ this threshold
Multi-label mask	Yes: Create single mask with all labels; No: Create separate mask per label

---

Workflow

Direct Copy

1. Open the Volume To Mask tool.
2. Select target volume(s).
3. Set Mode to Direct copy.
4. Click Apply.
5. New mask objects are created in the Object Browser.

Extracting Labels from Segmentation

1. Open the Volume To Mask tool.
2. Select the pre-segmented volume.
3. Set Mode to Segmentation.
4. Set Lower threshold and Upper threshold to define the label range.
5. Choose Multi-label mask:
   • Yes: All labels in one mask object
   • No: Separate mask for each unique label value
6. Click Apply.

---

Use Cases

Converting AI Segmentation Results

AI segmentation tools (TotalSegmentator, nnU-Net, MONAI) often output volumes where each anatomical structure has a unique integer label:

1. Import the segmentation result volume.
2. Use Segmentation mode with threshold range covering all labels (e.g., 1–117).
3. Create a multi-label mask for unified analysis, or separate masks for individual structures.

Creating Masks from Thresholded Data

For volumes already processed with threshold operations:

1. Use Direct copy mode.
2. Non-zero voxels become mask regions.
3. Fast conversion without additional processing.

Extracting Specific Structures

To isolate specific labels from a multi-label volume:

1. Set threshold range to the specific label value (e.g., lower=5, upper=5).
2. Only voxels with that label are included.
3. Creates a mask for just that structure.

---

Multi-Label vs. Separate Masks

Option	Result	Use Case
Multi-label (Yes)	One mask with multiple label values	Unified segmentation analysis, export as single file
Separate masks (No)	Individual mask per label	Independent editing, per-structure analysis

info

Multi-label masks preserve the original label values from the volume. Separate masks each contain only their respective label as the foreground value.

---

Output

Created masks appear in the Object Browser under the Masks category:

• Direct copy: Named based on source volume
• Multi-label mask: Single mask with all extracted labels
• Separate masks: Multiple masks, typically named with label indices

---

Data Type Considerations

Direct Copy Behavior

Source Data Type	Mask Result
Unsigned 8-bit	Direct copy (identical values)
Signed/Unsigned 16-bit	Rescaled to 0–255
32-bit integer	Rescaled to 0–255
Float/Double	Rescaled to 0–255

warning

Rescaling may lose precision for non-8-bit data. If exact values are important, use Segmentation mode with appropriate thresholds.

Segmentation Mode

Segmentation mode preserves label values within the mask (for multi-label) or uses the label value as the foreground (for separate masks).

---

Scripting

The Volume To Mask operations are available via the Python scripting API.

import ScriptingApi as api

app = api.Application()
volume_operations = app.get_volume_operations()

# Direct copy
mask_names = volume_operations.create_mask_by_direct_copying(["Volume1"])

# Segmentation - multi-label mask
mask_names = volume_operations.create_masks_by_segmentation(
    ["SegmentedVolume"],
    1,      # lower_threshold
    255,    # upper_threshold
    True    # create_multilabel_mask
)

# Segmentation - separate masks per label
mask_names = volume_operations.create_masks_by_segmentation(
    ["SegmentedVolume"],
    1,      # lower_threshold
    117,    # upper_threshold
    False   # create_multilabel_mask
)

See the VolumeOperations API Reference for details.

---

Related Tools

Tool	Direction	Description
Volume To Mask	Volume → Mask	Convert volume data to masks
Mask To Surface	Mask → Surface	Generate surface meshes from masks
Thresholding	Volume → Mask	Create masks using interactive thresholding