Workspace and Data

Workspace and Data

MedICS keeps in-memory data in a shared workspace. Python editors, Jupyter, toolboxes, and extensions all see the same variables.

---

Workspace model

• The workspace is a nested dictionary-like structure (DataDict).
• Each top-level key is a variable name (for example oct_volume, labels).
• Values are typically NumPy arrays, nested dicts, or MedImage objects.
• Updates emit a workspace_updated event so docks and views refresh.

Save / load

Action	Result
Save Workspace	Writes variables to a .med file (HDF5)
Load Workspace	Restores variables from .med
Auto-save	Optional periodic save (see [Application] auto_save_workspace)

---

Opening folders

File → Open Folder sets the current project folder used by:

• File Explorer
• Relative paths in scripts
• Some extension workflows

---

Supported image / data formats

Via FileIO and built-in toolboxes:

Family	Examples
DICOM	Including JPEG 2000 / OCT series
Volumes	TIFF stacks, NIfTI (.nii / .nii.gz)
MedICS	.med (HDF5 workspace / data), .medimage bundles
Scientific	MATLAB .mat, NumPy, HDF5
Images	PNG, JPEG
Video	.mp4, .avi, .mkv, .mov, .webm, …

---

Importing data

Use the ImportData toolbox:

1. Open Toolboxes → ImportData (or the activity-bar icon).
2. Drag files/folders, or browse a DICOM series.
3. Preview loads in the background with progress.
4. Confirm import — arrays appear under Variables and in Jupyter.

See Toolboxes.

---

Previewing files

Double-click in Explorer or use FilePreview:

• Text, markdown, code
• Tables / spreadsheets
• 2-D images
• Volumes (slice navigation, window/level)

---

Working from Python

In PyEditor or Jupyter, workspace variables are available in the shared namespace. Prefer wrapping volumes as MedImage when you need geometry, annotations, or provenance:

from medics.core.medimage import MedImage, SpatialGeometry, ImageMetadata

image = MedImage.from_numpy(
    volume,
    dims=("bscan", "depth", "aline"),
    geometry=SpatialGeometry(
        spatial_dims=("bscan", "depth", "aline"),
        spacing=(0.0468, 0.0039, 0.0117),
        coordinate_system="LPS",
        units=("mm", "mm", "mm"),
    ),
    metadata=ImageMetadata(modality="OCT"),
)

Legacy bare arrays still work through compatibility helpers (from_legacy_array / to_legacy_array).

---

Related pages

• MedImage
• Events and Data Flow
• Configuration