ImFusion Labels

Improve your workflow for image annotation

ImFusion Labels is designed to ease the workflow of medical image annotation for further algorithm development. In particular, it provides a user-friendly way of managing a database of images and annotating them in a few clicks.

A free demo is available here. Ready to unlock the full potential of ImFusion Labels? Click here to visit our webshop.

Database Management

  • Import all your images into a single project
  • Smoothly browse and visualize all datasets with associated labels
  • Search and sort your database with tags
  • Use the server version for remote and concurrent data annotation

Image Annotation

  • Smart algorithms to annotate images, volumes, and sequences
  • Fast multi-label segmentation and post-processing tools
  • Import and refine existing annotations
  • Write your own interactive algorithms in C++ or Python

Customization

  • Extend the functionalities of the software with custom plugins
  • Example: The Spine Plugin includes a new labeling mode, dedicated to spine, that speeds-up the annotation process

Experiments

  • Try your algorithms on the whole dataset and compare results to the ground truth

Standardized Export

  • Export your database directly for your training pipeline
  • Choose the file format, encoding, resolution, etc.
  • Add data augmentation to generate more images

Key advantages over existing solutions

  • the support of a large variety of image modalities and data format, including DICOM images,
  • a toolset of segmentation algorithms that have been designed to be as fast and intuitive as possible,
  • a Python integration, which allows user to write their own algorithms to help them label data or run experiments on all existing ones,
  • a powerful and customizable visualization of data and their labels,
  • the possibility to easily define post-processing (resampling, orientation normalization, data augmentation, etc.) before exporting the database.