Recent Releases

ITK 4.9 Released

The Insight Segmentation and Registration Toolkit (ITK) 4.9 release adds new remote modules, such as Cuberille for smooth generation of meshes from binary segmentation images. ITK 4.9 also enhances the edge-preserving smoothing abilities of the AnisotropicDiffusionLBR module.

In addition, itk::Image::GetPixel performs better, as do operations on the itk::VariableLengthVector class. Plus, Python wrapping builds faster with a more standard installation.

The 4.9 release makes it possible to build an ITK module outside the source tree. The release also enables adding continuous integration testing on GitHub, as well as Python wrapping for C++ algorithms. In other news, ITK now generates a FlatStructuringElement from an image. ITK also completes point set registration with time-varying B-splines and specifies module dependencies as public / private / compile / test.

On behalf of the ITK community, congratulations and thank you to everyone who contributed to the release. For more information and to download the release, please visit

A cuberille filter generates meshes based on images, as described in “Cuberille Implicit Surface Polygonization for ITK” on (Insight Journal/Dan Mueller)

ParaView 5.0 Released

ParaView 5.0 introduces several changes to the open-source, multi-platform data analysis and visualization application. These changes concern ParaView’s rendering engine, OpenSWR and Mesa support, OSPRay integration, off-screen rendering, in situ analysis, and accelerated algorithms.

Rendering Engine
ParaView has a brand-new backend that implements the redesigned rendering engine of the Visualization Toolkit (VTK). The rewrite of the engine began in 2014 with the aim of adopting modern OpenGL application program interfaces (APIs) and design patterns. Targeted for modern graphics processing units (GPUs), the new engine has almost all of the capabilities of the old engine, along with significant performance improvements for both polygonal and volumetric rendering. Binaries distributed on ParaView’s website have VTK’s new rendering engine enabled.

OpenSWR and Mesa Support
ParaView Linux binaries now include the initial public release (alpha) of Intel’s Mesa and OpenSWR integration work. OpenSWR is a software rasterizer developed by Intel. It runs on laptops, workstations, and compute nodes in high-performance computing (HPC) systems.

OSPRay Integration
ParaView Linux binaries also include the pvOSPray plug-in. The OSPRay Rendered 3D View makes additional display and view properties available for controlling parameters such as shadows. OSPRay is an open-source ray-tracing engine developed by Intel.

Off-screen Rendering
The Embedded-System Graphics Library (EGL) Native Platform Interface provides means to render to an off-screen buffer or a native windowing system, such as Android, X Windows, or Microsoft Windows. ParaView command line executables (e.g., pvserver, pvbatch, and pvpython) can now render to an EGL off-screen buffer. ParaView 5.0 also allows for server-side, hardware-accelerated rendering without installation of a windowing system.

In Situ Analysis
The 5.0 release of ParaView provides a Community Atmosphere Model version 5 (CAM5) ParaView Catalyst adaptor, enabling in situ analysis for CAM5 simulations. ParaView Catalyst is a lightweight version of the ParaView server that embeds directly into parallel simulations to perform in situ analysis at run time.

Accelerated Algorithms
Highly scalable for large data, the new AccleratedAlgorithms plug-in includes FlyingEdges2D and FlyingEdges3D algorithms, which generate isosurfaces from two-dimensional (2D) and three-dimensional (3D) data, respectively. The plug-in also features FlyingEdgesPlaneCutter. This algorithm cuts a volume with a single plane. For more information, please read the paper “Flying Edges: A High-Performance Scalable Isocontouring Algorithm” by Will Schroeder, Robert Maynard, and Berk Geveci.

Try it Out
To learn about leveraging ParaView, please visit ParaView’s website at To try the latest release, please go to

3D Slicer 4.5 Released

3D Slicer 4.5 supports basic and applied medical imaging research. 3D Slicer is an industry-leading software application for visualizing and processing medical images from a multitude of medical devices. Researchers use it to correlate the shape of structures in magnetic resonance imaging (MRI) scans of brains with the risk of schizophrenia, as well as to measure cardiac function from computed tomography (CT).

A customized version of 3D Slicer, which is part of a surgical planning system to enable optimal cranial remodeling in children with craniosynostosis, generates a multi-level segmentation of a child’s skull bones.

One of the many strengths of 3D Slicer is its extensible, open-source foundation. The 4.5 release offers over 60 extensions and 150 feature improvements, in addition to performance and stability advancements. A community of over 160 researchers from around the world provide the extensions for free academic and commercial use. These extensions enable 3D Slicer to import numerous new kinds of medical images and to make precise measurements of a variety of anatomic structures and pathologies. One extension, for example, computes biomarkers for lung disease from CT images.

The 4.5 release tightens the integration between 3D Slicer and the Insight Segmentation and Registration Toolkit (ITK). It also takes advantage of recent developments in the Visualization Toolkit (VTK). These toolkits, like 3D Slicer, contain best-in-class scientific methods and employ the high-quality software practices of Kitware, based on CMake, for building and distributing stable applications on Windows, Linux, and Mac OS platforms.
For a full set of features of 3D Slicer 4.5, please visit the Wiki page of 3D Slicer. To learn more about 3D Slicer and its applications in medical imaging, please visit

CMake 3.4 Released

CMake 3.4 introduces new capabilities to the open-source, cross-platform build system to further enhance the way developers build, test, and package software.

Several new capabilities added to CMake in version 3.4 involve commands. The if() command, for example, knows a new TEST operator that evaluates to true if the add_test() command defines a given test name. Also notable, the DESTINATION option for install(DIRECTORY) and install(FILES) commands knows to support generator expressions.

Moreover, CMake now honors *.manifest source files with Microsoft Visual C++ (MSVC) tools. CMake will merge manifest files named as sources of .exe and .dll targets with linker-generated manifests and embed them in the binary.

Version 3.4 also includes improvements to generators, variables, properties, modules, and generator expressions. For full release notes and to download CMake, please visit

Girder 1.4 Released

Girder 1.4 improves install utilities to make it easier to install third-party plug-ins. For example, the updated girder-install command and general packaging infrastructure simplify the process of bundling and installing third-party plug-ins that contain Web client code.

Version 1.4 also simplifies the Grunt build process, thanks to refactored Grunt tasks. The release refactors the OAuth plug-in too, making it possible to support new providers such as GitHub and to enable a subset of providers with less effort.

Furthermore, Girder now features a system setting for the creation of default Public and Private folders for new users. Also new, the Girder python client instantiates GirderClient with a full URL rather than requiring users to specify the URL in parts. Additionally, an added Representational State Transfer (REST) endpoint deletes the contents of a folder without deleting the folder itself.

Girder is a free, open-source, Web-based data management platform that is part of Resonant data and analytics ecosystem. To learn more about Girder, please visit

CMB 4.0 Released

Computational Model Builder (CMB) 4.0 can interact with discrete geometry in Exodus, Network Common Data Form (NetCDF), the Visualization Toolkit (VTK), and Moab. The release includes full simulation attribute support, user interface customization through eXtensible Markup Language (XML), and export capabilities via Python.

CMB is a framework and application suite that enables various toolkits and modeling systems to form an end-to-end solution for supporting simulation life cycles. Kitware develops CMB open source and offers it with a permissive, three-clause Berkeley Software Distribution (BSD) license. For more information and to download CMB, please visit

This image shows a model of an Advance Burner Test Reactor core loaded in CMB’s ModelBuilder application.

Questions or comments are always welcome!