The Medical Image Segmentation Tool Set iSEG, originally developed by the IT'IS Foundation as an in-house solution for the creation of computational anatomical phantoms such as its widely used Virtual Population (ViP), is now open source. This important step will facilitate the further improvement, extension, and eventual re-purposing of the iSEG platform for communities beyond its original scope.
The iSEG medical-image segmentation and modeling toolkit is now open source. Watch the introductory tutorial here.
Compared to alternative software, iSEG is able to handle enormous 3D image datasets and manage large numbers of different tissues or complex tissue hierarchies, making it the ideal toolbox for the efficient, fast, and flexible generation of anatomical models from various types of imaging data. Ever since its development, iSEG has been integrated into our Sim4Life simulation platform as an image-modeling module that allows users to create patient-specific anatomical geometries for dosimetry, hyperthermia, neuro-stimulation, and other life science applications.
The decision to make iSEG open source was made in recognition of the multitude of diverse questions that arise in medical-image segmentation, with powerful problem-specific algorithms and workflows. iSEG is released under the very permissive MIT license on GitHub, which enables users to create their own extensions. To support this feature, we implemented an easy-to-use plugin mechanism to allow the addition of new algorithms with simple user interfaces without having to modify the iSEG core.
We hope this approach will encourage a growing community of researchers and developers to use and extend iSEG for their individual projects. Instructions for building from the source code and tutorials for first-time users are available on the IT’IS GitHub page.
ZMT and IT’IS remain committed to maintaining and extending iSEG in the future.