October 2012 Kitware News – Part One

October 9, 2012

Kitware Presents Python Tutorial at CVPR
Over the summer at the CVPR conference, Kitware collaborated with Google Research to present the “Python for MATLAB Users” tutorial. The tutorial was taught by Matt Leotta, Eran Swears, and Patrick Reynolds from Kitware, and Yong Zhao and Varun Ganapathi from Google Research.

The course provided a hands-on introduction to using Python for rapid prototyping and computer vision research. Python was presented from a Matlab users’ point-of-view, showing how to directly map the familiar Matlab syntax and workflow to a very similar syntax and workflow in Python. The team demonstrated a free and open-source alternative to Matlab for promoting open science, and to better prepare students for careers outside of academia.

Attendees received a virtual appliance (VirtualBox) with the full scientific Python environment pre-configured, and executed programming exercises ranging from introductory to intermediate levels. The material, including the virtual appliance, was distributed on pre-configured USB memory sticks and DVDs. Participants spent the day working on exercises, asking questions, and collaborating.

Based on the positive feedback from participants and CVPR attendees who wanted to but could not join the tutorial, the material has been made available on Kitware’s website at the tutorial page: http://www.kitware.com/cvpr2012.html.

Avogadro in Journal of Cheminformatics
“Avogadro: An advanced semantic chemical editor, visualization and analysis platform” was recently published in Volume 4 of the Journal of Cheminformatics. The paper was authored by Marcus D Hanwell, Donald E Curtis, David C Lonie, Tim Vandermeersch, Eva Zurek, and Geoffrey R Hutchinson, and discusses Avogadro 1.0.

Avogadro, the open-source, cross-platform molecule editor and visualization platform is designed for use in computational chemistry, molecular modeling, bioinformatics, materials science, and related research areas. The paper describes the platform’s direct applications in research and education as they relate to chemistry, physics, materials science, and biology, and includes examples on how to extend the tool via plugins. The Avogadro platform is being developed further as part of the Open Chemistry initiative at Kitware.

ITK Chapter in AOSA Book Volume II
Last year, four Kitware authors contributed two chapters, one on VTK and another on CMake, to “The Architecture of Open Source Applications.” In the next edition, “The Architecture of Open Source, Volume II: Structure, Scale, and a Few More Fearless Hacks,” Luis Ibáñez and Brad King contributed a chapter on the Insight Segmentation and Registration Toolkit (ITK). This chapter covers what ITK is, its architectural features, and lessons learned during development.

The book is edited by Greg Wilson and Amy Brown, and features chapters on many other open-source projects.  The book is available online or as a printed copy through Lulu and is distributed under the Creative Commons BY license.

Kitware #1245 on the 2012 Inc. 5000 List
In August, Inc. Magazine announced its 2012 Inc. 500|5000 list, an exclusive ranking of the nation’s fastest-growing private companies. Kitware ranked #1245 overall, with three year revenue growth of 248%.

This is the fifth year that Kitware has been featured on the Inc 5000 list, and is Kitware’s highest rank to date! With an industry ranking of #100, Kitware attributes its success to its open-source business model and support of talented employees who are energized about the company’s research initiatives and mission.

The ParaView Guide Published as an eBook
The ParaView Guide was released as an eBook this summer, and is now available from Kitware’s website. This edition of the ParaView Guide has been updated for ParaView 3.14 and is Kitware’s first foray into the world of electronic publishing. The content of the revised ParaView Guide comes from the ParaView wiki, with additional chapters covering CFD post-processing, an introduction to visualization with ParaView, and visualization and analysis of astrophysical AMR datasets. As an eBook, Kitware will be able to update and disseminate new versions of the Guide more quickly and efficiently to keep pace with real-world, digital demand. Additionally, users who purchase the eBook can upload and access it from any number of their electronic devices.

If you have any feedback about the eBook or topics you’d like to see in future editions, please let us know by emailing editor@kitware.com.

Kitware Wins NIH Phase II SBIR Award
In August, Kitware announced a new Phase II SBIR grant from the National Institutes of Health to develop a computational infrastructure for mapping the mammalian connectome. The nearly $1M project is a collaborative effort between Kitware and Harvard University.

A human brain is estimated to have roughly 100 billion neurons connected through more than 100 thousand miles of axons and a quadrillion synaptic connections. The neural circuits within each brain are its connectome; understanding how the connectome works to enable cognition, consciousness, and intelligence is one of the most fundamental questions in science.
Advances in modern technology have made it easier to collect and prepare tissue samples for high-resolution scanning. This leads to digital representations of the connectome that are currently 60-70 terabytes in size, with this expected to grow rapidly as imaging technology improves.

Charles Law, Principal Investigator, will lead the development of a computational infrastructure that will include a set of segmentation algorithms that can extract neurons and their synapses from electron micrographs of serial tissue sections, and support arbitrarily large image volumes that will allow researchers to trace neural processes. Additionally, the suite will include fusion methods that take segmentation results as input and generate tracings of neural processes by linking segmentation results from one section to the next.

Kitware Wins Phase II Funding to Further Develop ClimatePipes
Kitware was awarded nearly $1M in Phase II funding from the U.S. Department of Energy to further develop ClimatePipes, a platform for providing non-researchers with access to and analysis tools for understanding long-term climate change projections. This is a collaborative project with researchers at the Polytechnic Institute of New York University.

The pace of global climate change is expected to accelerate and impact a variety of fields, including public health, infrastructure, and water resources. In recent years, multitudes of climate data have been collected as part of the U.S. Global Change Research Program and the Coupled Model Intercomparison Project (CMIP); however, hurdles including large data size, a lack of adequate metadata, poor documentation, and insufficient computational and visualization resources have limited research progress in this area.

Upon completion, ClimatePipes will make such valuable climate data available to non-researchers including policy makers, health officials, agriculturists, and industry leaders. Users will be able to run complex computations and visualizations in the cloud or on a user-provided cluster, and upload data to a server for analysis and integration.

In Phase I, ClimatePipes was developed as a web-based workflow with interfaces for accessing, querying, and visualizing datasets from one-or-more sources. In Phase II, Kitware will implement support mechanisms for more elaborate and relevant queries, and improve the system’s usability, robustness, and scalability. Advanced querying will be supported by a semantic search tool that relies on natural language processing techniques.

This work has been supported by the Department of Energy (DOE) under award number DE-SC0006493.

Keep reading October 2012 Kitware News – Part Two

Leave a Reply