nitrain/nitrain

Health v2 · maintenance only

Adoption not part of the score

1880 stars · 299 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Nitrain is a framework-agnostic Python library for sampling, augmenting, and training AI models on medical imaging datasets, with support for PyTorch, TensorFlow, and Keras backends. It provides high-level abstractions for datasets, loaders, transforms, and trainers tailored to medical images like NIfTI files.

Use cases

• train a segmentation model on brain MRI scans
• augment medical images for deep learning
• build a dataset from NIfTI files with transforms
• train models on medical images with pytorch or keras
• sample 2D slices from 3D medical image volumes
• visualize medical imaging model results

When to choose

• you need a high-level, framework-agnostic API for medical imaging AI
• you work with NIfTI/neuroimaging formats and want built-in transforms and samplers
• you want to switch between torch, tensorflow, and keras backends

When to avoid

• you train models on natural images rather than medical imaging formats
• you need fine-grained low-level control over data pipelines
• AGPL-3.0 licensing is incompatible with your project

Facets

library · maturity active

machine-learning deep-learning image-processing data-science machine-learning deep-learning healthcare python medical-imaging neuroimaging pytorch keras tensorflow data-augmentation medical-imaging-ai antspy

Member repositories

For agents

markdown · JSON · MCP: product_card(name="nitrain/nitrain")