ndpi2bids 0.1.2

Convert Hamamatsu NDPI whole-slide images to BIDS-compliant microscopy datasets

BIDS microscopy: ndpi

---

Convert whole-slide images in .ndpi format to Brain Imaging Data Structure (BIDS) microscopy compliant datasets. https://hub.docker.com/r/micalab/ndpi2bids

Overview

BIDS_ndpi automates the conversion of Hamamatsu .ndpi whole-slide microscopy images into a BIDS-compliant directory structure, generating the required JSON sidecars, participants.tsv, and metadata files along the way.

---

Repository structure

BIDS_ndpi/
├── CHANGELOG.md
├── Dockerfile
├── LICENSE
├── README.md
├── boutiques
│   └── ndpi2bids.json                  # Boutique descriptor 
├── environment.yml
├── .gitignore
├── ndpi2bids
│   ├── __init__.py
│   ├── ndpi2bids.py                    # Main conversion script
│   └── templates
│       ├── dataset_description.json
│       ├── participants.json
│       ├── stain-AT8_BF.json           # Stain-specific JSON sidecar template
│       └── CITATION.cff
└── pyproject.toml

---

Requirements

• Docker (recommended), or
• conda with the provided environment.yml
• pip

---

Quick start

With Docker

# Build the image
docker build -t bids_ndpi .

# Run conversion
docker run --rm \
  -v /path/to/ndpi/files:/input \
  -v /path/to/output:/output \
  bids_ndpi \
  python ndpi2bids/ndpi2bids.py --input /input --output /output

With conda

# Create environment
conda env create -f environment.yml
conda activate bids_ndpi

# Run conversion
python ndpi2bids/ndpi2bids.py --input /path/to/ndpi --output /path/to/bids

---

Usage

python ndpi2bids/ndpi2bids.py [OPTIONS]

Mandatory arguments:
  --ndpi_path  PATH   Path to raw Hamamatsu .ndpi file             [required]
  --bids       PATH   Path to the root of the BIDS dataset         [required]
  --sub        LABEL  Subject ID (e.g., PX067)                     [required]
  --stain      LABEL  Stain entity (e.g., AT8)                     [required]
  --suffix     LABEL  BIDS suffix (e.g., BF)                       [required]

Optional BIDS entities:
  --ses        LABEL  Session ID
  --sample     LABEL  Sample ID (e.g., NP24709)
  --acq        LABEL  Acquisition label
  --run        INDEX  Run index
  --chunk      LABEL  Chunk label (e.g., A3)
  --template   PATH   JSON template for metadata (overrides defaults)

Operational flags:
  --convert           Convert NDPI to OME-TIFF via bfconvert
  --force             Overwrite existing BIDS files and sidecars
  --dry_run           Print the intendet output.

Example

python ndpi2bids/ndpi2bids.py \
  --ndpi_path /data/raw/PX067_AT8.ndpi \
  --bids      /data/bids \
  --sub       S001 \
  --stain     AT8 \
  --suffix    BF \
  --ses       01 \
  --sample    NP24709A1 \
  --convert

---

Templates

JSON templates in ./templates/ define stain-specific metadata injected into BIDS sidecars.

File	Purpose
stain-AT8_BF.json	Sidecar template for AT8 brightfield staining
participants.json	Column definitions for participants.tsv
participants_description.json	Dataset-level participant metadata
CITATION.cff	Citation metadata for the dataset (authors, title, version, DOI)

To add a new stain, copy an existing template and update the fields, then pass it via --template.

---

BIDS output structure

output/
├── CITATION.cff
├── dataset_description.json
├── participants.tsv
├── participants.json
├── README
└── sub-<label>/
    └── ses-<label>/
        └── micr/
            ├── sub-<label>_ses-<label>_stain-AT8_BF.ome.tif
            └── sub-<label>_ses-<label>_stain-AT8_BF.json

---

License

MIT License — see LICENSE for details.