Medical Imaging Pipelines

What This Skill Does

Generates code for end-to-end medical imaging data pipelines: ingestion, format conversion, preprocessing, dataset preparation, and export. Covers the common path from raw DICOM files to ML-ready datasets, research exports, and automated processing workflows.

Prerequisites

• Python 3.8+
• Core: pydicom, numpy
• Conversion: SimpleITK or nibabel (for NIfTI), pillow (for PNG/JPEG)
• ML prep: scikit-image, scipy
• Optional: h5py (HDF5), pandas (metadata), tqdm (progress bars)

# Full pipeline toolkit
pip install pydicom numpy SimpleITK nibabel pillow scikit-image scipy h5py pandas tqdm pylibjpeg pylibjpeg-libjpeg