# TorchIO-project/torchio

Medical imaging processing for AI applications.

Repository: https://github.com/TorchIO-project/torchio
Canonical: https://ross.abutalabs.com/products/torchio
Homepage: https://docs.torchio.org/
Language: Python
License: Apache-2.0
License Family: permissive
Topics: pytorch, medical-image-computing, deep-learning, data-augmentation, machine-learning, python, medical-image-processing, medical-image-analysis, augmentation
Last push: 2026-08-01T19:06:40+00:00

## Health v2 (maintenance only)
Score: 93/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 95, release rhythm 86, longevity 100
- inputs: {"age_days": 2472, "days_push": 32, "days_rel": 92, "gap_med": 0, "n_releases_24m": 52}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2439, forks 275 (observed 2026-08-28T04:06:51.587599+00:00)

## What it is
TorchIO is a Python library for loading, augmenting, and processing 3D medical images (MRI, CT) within PyTorch deep learning pipelines. It provides medical-image-specific transforms, samplers, and dataset utilities for training AI models.

## Use cases
- augment 3D MRI scans for training a segmentation network
- load NIfTI medical images into PyTorch datasets
- train a deep learning model on CT volumes
- apply spatial and intensity transforms to medical images
- preprocess medical imaging data for AI research

## When to choose
- you are doing deep learning on 3D medical images with PyTorch
- you need medical-specific augmentation like random bias field or elastic deformation
- you want efficient patch-based sampling of large volumetric scans

## When to avoid
- you work with 2D natural images rather than volumetric medical data
- you use TensorFlow or another framework instead of PyTorch
- you only need general-purpose image viewing or DICOM conversion without training

## Facets
- artifact type: library
- maturity: active
- function: image-processing, machine-learning, deep-learning, data-generation
- domain: machine-learning, deep-learning, healthcare, image-processing
- platform: python, cross-platform
- tags: pytorch, medical-imaging, data-augmentation, mri, ct, nifti, 3d-images, gpu

## Member repositories
- TorchIO-project/torchio (main) score 93

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:51.587599+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:30:45.721108+00:00, confidence not recorded.
  - readme: https://github.com/TorchIO-project/torchio (fetched 2026-08-28T04:06:51.587599+00:00, sha 8bbccfaf4167)
  - homepage: https://docs.torchio.org/ (fetched 2026-08-29T10:12:14.151510+00:00, sha 36c6c3c2e4f9)
- Data as of 2026-08-30T08:39:29.467469+00:00.