Description

The imfusion package enables easy and fast loading, handling and processing of medical image data. It is a wrapper around the ImFusion SDK and exposes a subset of its functionality to Python. The major advantages of using imfusion are:

• High Performance:

  Leveraging optimized C++ for fast execution and OpenGL for GPU acceleration, ensuring compatibility with various GPU vendors.

• Versatile Data Structures:

  Handle a wide range of medical images and data types, including 2D/3D images, metadata, deformations, rotations, masks, and segmentations. It also supports keypoints, point clouds, and meshes.

• Extensive Set of Algorithms:

  Access a vast array of image processing algorithms, from basic cropping to complex multi-modal image registration. Even algorithms that don’t have dedicated Python bindings can be executed through a functional interface.

• File Format Support:

  Load and save numerous medical imaging formats, including Nifti, MHD, Dicom, HDF5, PNG, and JPG and featuring a reliable Dicom loader used in FDA-approved products.

• Deployment-Ready Data Pipelines:

  Construct efficient data pipelines for ML model training and deployment, ensuring consistent pre-processing and post-processing.

• numpy-like arithmetic but with images

  Perform arithmetic operations on images with a functional API or operators, supporting GPU or CPU execution and, optionally, aggregated expression evaluation for enhanced performance.

• Built-in Data Visualization

  View your data with the bundled ImFusionVisualizer (imfusion.show) that includes a powerful 3D renderer for volumes, meshes and point clouds.

Documentation

Please find the documentation for this Python package at docs.imfusion.com/python.

Support

If you experience issues with this package, please let us know in our forum.

For business inquiries please contact info@imfusion.com.