pentachrome-plugin 0.4.0

Napari plugin for the Pentachrome histology pipeline: VSI extraction, nnUNet inference, statistics.

pentachrome-plugin

Napari plugin for the Pentachrome histology pipeline. Current widgets:

• VSI to TIFF Extractor (Phase 1) — extract tissue-region TIFFs from VSI files.
• nnUNet Inference (Phase 2) — run the trained Epithelium / MultiStructure models on selected TIFF layers and load colorized masks back into the viewer.
• Mask Statistics (Phase 3) — per-region statistics (thickness, composition, cell densities) computed on the inference output, with CSV export.

Source and issues: https://github.com/dtsilis7/PentachromePipeline

Requirements: Windows, Python 3.10, napari >= 0.4.18, and a Java JDK (JDK 17 confirmed). The Java/bioformats stack and the nnUNet model weights are installed separately (see below) — they can't come from a plain pip install.

Install (Windows, PowerShell)

Requires a working Java JDK on PATH (JDK 17 confirmed working).

Via napari plugin manager (after publishing)

Once this package is published to PyPI it shows up in napari's Plugins -> Install/Uninstall Plugins (search "pentachrome"). That installs the Python package only — you still need the conda-forge Java/bioformats step and the nnUNet weights below for the extractor and inference to work.

Recommended: conda env

cd into the plugin directory first — pip install -e . resolves . relative to your current shell directory:

conda activate napari
cd "...\pentachrome_plugin"
conda install -c conda-forge python-javabridge python-bioformats
pip install opencv-python tifffile
pip install -e .

If you would rather not cd, pass the absolute path explicitly:

pip install -e "...\pentachrome_plugin"

Conda-forge ships pre-built wheels for python-javabridge and python-bioformats and avoids the MSVC + NumPy-2 compile failure that pip install python-javabridge hits today (the C extension references _PyArray_Descr fields removed in NumPy 2.0).

Fallback: pure pip

Only use this if conda-forge is unavailable. NumPy must be pinned below 2 before javabridge builds, and build isolation must be off so the build sees the pinned NumPy:

cd "...\pentachrome_plugin"
pip install "numpy<2"
pip install --no-build-isolation python-javabridge python-bioformats
pip install opencv-python tifffile
pip install -e .

Launch

conda activate napari   # or whichever env you installed into
python -m napari

In napari: Plugins -> VSI to TIFF Extractor or Plugins -> nnUNet Inference.

nnUNet Inference (Phase 2)

Requires nnunetv2 installed in the same environment (the nnUNetv2_predict CLI must be on PATH).

conda activate napari
pip install nnunetv2

Workflow:

1. Load TIFFs into napari (e.g. via Phase 1's auto-load checkbox, or drag-and-drop).
2. Open Plugins -> nnUNet Inference.
3. Select one or more image layers in the list.
4. Tick Epithelium, MultiStructure, or both.
5. Set Output folder (where raw + colorized masks go) and nnUNet results (folder containing Dataset001_Epithelium and Dataset002_MultiStructure). The results path auto-fills if nnUNet_Training/nnUNet_results/results is found.
6. Pick Device (cpu or cuda) and click Analyze.

Speed vs quality (important on CPU)

nnUNet inference on a laptop CPU is slow because every image goes through folds × mirror augmentations × sliding-window patches forward passes. With defaults that can be 20+ passes per image. The widget exposes three knobs in the Speed / quality group:

Knob	Default	What it does
Epithelium folds	Fold 0 only	Use 1 of the 5 trained folds for Dataset001. All 5 ensembled is best quality but ~5x slower. Dataset002 only has fold 0 trained, so it's always 1 fold.
Disable test-time mirroring	on	Passes --disable_tta. Skips the 4 mirror augmentations the model normally averages over. ~4x faster, small accuracy hit.
Sliding-window step	0.5	Passes -step_size. Larger = fewer overlapping patches = faster but rougher tile borders. Try 0.7 for a middle ground.

With all three defaults on a CPU laptop, one ROI tile should take a few minutes instead of 30+. Switch to All 5 folds + TTA on once you've moved to a GPU box.

Continuing from the extractor

The two widgets are linked through two small bridges, so you can run Extract -> Analyze in a single napari session without re-picking files:

• When the extractor auto-loads a TIFF as a viewer layer, it stashes the on-disk path on layer.metadata['source_tiff']. The inference widget reads that during staging and copies the original file into _staging_input/ rather than re-saving the in-memory array — important for 15k x 15k tiles.
• When an extraction completes, the inference widget's "Use last extractor output" button pre-fills the output folder to <extractor_output_root>/_inference, so masks land next to the per-VSI subfolders the extractor created.

Both bridges are in-process only (see _session.py); they reset when napari closes.

Outputs land in:

<output_folder>/
  _staging_input/            # nnUNet-named (_0000.tif) copies of the selected layers
  epithelium_raw/            # binary masks from Dataset001
  epithelium_colored/        # RGB colorized masks (red epithelium)
  multistructure_raw/        # 6-class masks from Dataset002
  multistructure_colored/    # RGB colorized masks (Elastin/Collagen/Nuclei/Mucins/Membrane/Goblets)

Colorized masks are added to the viewer as RGB image layers when the run finishes.

nnUNet inference architecture

Same subprocess pattern as Phase 1. The widget never imports torch or nnUNetv2 directly; it spawns _inference_worker.py which:

• sets nnUNet_results to the configured results dir,
• calls nnUNetv2_predict once per enabled model (folds 0-4 for Epithelium, fold 0 for MultiStructure, matching run_inference.py),
• colorizes the resulting integer masks with the palettes from colorize_masks.py / compare_grid.py,
• streams JSON-line events on stdout for the widget's progress bar and log.

How it works

• The widget itself never touches the JVM. When you click Extract ROIs, it spawns _vsi_worker.py as a separate Python process.
• That worker process starts the bioformats JVM, loops over the VSI files using TileMaskStitcher (reused from VSI_Handler/tile_mask_stitcher.py), writes numbered TIFFs into <output_root>/<vsi_basename>/, and emits JSON-line progress events on stdout.
• The widget streams those events on a background thread and updates the progress bar / log without blocking the UI.
• When the worker exits, the JVM dies with it. The next extraction batch starts a fresh JVM in a fresh process - this avoids the "JVM cannot be restarted" pitfall during a long napari session.

Defaults

The parameter defaults mirror Processing_VSI_Files.py:

Parameter	Default
Series	6
Tile width / height	15000
Threshold	50
Min ROI area	150000
Merge margin	1000
Extra crop margin	100

Layout

pentachrome_plugin/
  pyproject.toml
  README.md
  src/pentachrome_plugin/
    __init__.py
    napari.yaml             # napari manifest
    _session.py             # in-process cross-widget state (extractor -> inference -> analysis)
    _widget.py              # VsiExtractorWidget (Phase 1)
    _vsi_worker.py          # VSI subprocess entrypoint
    _inference_widget.py    # NnUnetInferenceWidget (Phase 2)
    _inference_worker.py    # nnUNet subprocess entrypoint
    _analysis_widget.py     # AnalysisWidget (Phase 3, in-process)

Phase 3 (Mask Statistics) lives alongside these and registers through napari.yaml.

Mask Statistics (Phase 3)

Pure in-process; no subprocess needed (no JVM, no torch). Reuses EpithelialAnalysis/Analyzers/ (Descriptors.py, Thickness.py), so the same metrics that fed the original region_summary.csv show up in the widget.

Workflow:

1. Run Phase 2 first so epithelium_raw/ and multistructure_raw/ exist.
2. Open Plugins -> Mask Statistics.
3. Select one or more image layers in the list (their names must match the mask filenames in epithelium_raw/ / multistructure_raw/; if the inference widget staged them, that's already true).
4. Click Use last inference output (or browse).
5. Tweak Pixel size, Region dilation, Min epithelium area if needed (defaults match Main.py).
6. Click Analyze.

For each detected epithelial region the widget reports:

Column	What it is
Area (mm^2)	Region area after the 50 um dilation
Thickness mean/std (um)	Medial-axis thickness of (membrane within eroded region) U goblets U nuclei
Elastin / Collagen / Other %	Fraction of stained structure pixels — same definition as compute_structure_percentages
Mucin %	Mucin pixels as a fraction of the epithelium area (not of total structure pixels)
Nuclei / mm^2 and Goblets / mm^2	Density per mm^2 of epithelium — goblet hyperplasia is a classic COPD readout
Nuclei (n), Goblets (n)	Raw counts inside the region

A bold (all regions) row appended per image gives area-weighted means of the percentages / thickness and totals for the counts. Export CSV... saves the whole table (per-region rows + aggregate rows).

The elastin organization score (ElastinAnalyzer.determine_organized_region) from Main.py is intentionally not yet exposed — it's much heavier (skan + shapely + ROI polygons) and will land as a separate toggle.

Class isolation

A "Class isolation" group at the top of the widget lets you view a single class (or a combination) without rerunning anything:

1. Pick a source layer (the original TIFF — not a colorized mask).
2. Tick one or more of Elastin, Collagen, Nuclei, Mucins, Cell Membrane, Goblets, Epithelium.
3. Click one of:
   • Show as mask — adds a new layer that's white everywhere except the ticked classes, colored with the same palette as the inference widget.
   • Show on original — adds a copy of the original image with all pixels outside the ticked classes turned white. Useful for sanity- checking the segmentation against the stain.
4. Clear isolated layers removes everything this panel added in one go.

Masks are read on demand from the inference output folder; the original layer's pixels are taken from the viewer.

License

MIT. (Matches license = {text = "MIT"} in pyproject.toml. Consider adding a top-level LICENSE file with the MIT text so it ships in the sdist/wheel.)