ITK/Release 4/Outreach/Conferences/MICCAI 2011/ITKv4

ITKv4: The Next Generation

Introduction

The Insight Segmentation and Registration Toolkit (ITK) has been the leading software resource used in the field of medical image analysis for more than ten years now. In order to cope with the continuously changing demands of this field, the US National Library of Medicine has invested in 2010-2011 close to five million dollars in revising, improving, simplifying and accelerating the toolkit, and improving its DICOM support. A team more than 50 developers from ten different institutions have collaborated in this demanding project. The outcome of this effort is ITKv4, and will be released shortly before the MICCAI conference.

• http://www.miccai2011.org/workshops

Goals

The purpose of this half-day tutorial is to introduce attendees to the new capabilities and features of this fresh version of ITK, and to provide guidance on how they can migrate their own applications to start taking advantage of the many improvements now available in ITKv4.

Target Audience

• Existing users of ITK who want to be familiarized with the new changes in ITKv4.

Relevance to MICCAI

The large community of ITK users will be interested in learning about the new features in ITKv4 that include:

• Support for GPU computation
• Improved image registration framework
• Improved level set framework
• Improved DICOM support
• Modularization of the toolkit

Content

• What is new in ITKv4 ?
  • Registration framework
  • Level Sets framework
  • GPU support
  • FEM framework
  • Improved DICOM support
  • Modularization
• How to migrate
  • Hands on exercises on migration
• Better support for microscopy image analysis

Educational Objectives

At the end of this tutorial, attendees will have gained practical experience on:

• Exercising the new functionalities of the image registration framework
• Exercising the new features of the level sets framework
• Exploiting the new capabilities of GPU computation
• Taking advantage of the new modular architecture of ITK source code distribution
• Using the resources available for supporting their migrating to ITKv4

Speakers

The speakers for this tutorial will be:

• Luis Ibanez, Ph.D. Technical Leader, Kitware Inc.
• Brian Avants, Ph.D., University of Pennsylvania
• Bradley Lowekamp, Software Engineer, NLM-NIH.
• Dan Blezek, Ph.D., Mayo Clinic
• Xiaoxiao Liu, Ph.D., R&D Engineer, Kitware Inc.
• Gabe Hart, M.Sc., R&D Engineer, Kitware Inc.

Presentations

• Overview of ITKv4 (30min) Luis Ibanez
• Installation of resources for hands-on exercises (Virtual Appliance) (30 min) Gabe Hart
• Level Sets new features (30 min) Luis Ibanez
• Image Registration new features (45 min) Brian Avants
• Break (15min)
• Modularization (30 min) Xiaoxiao Liu
• GPU computation support (30 min) Dan Blezek
• Microscopy Applications (30 min) Bradley Lowekamp

Methodology

This is a hands-on tutorial for image analysis practitioners who have previous experience with ITK. Attendees are required to bring their laptops to the tutorial. At the beginning of the tutorial, attendees will receive USB memory sticks with a full installation of a VirtualBox virtual appliance, containing a Ubuntu 10.10 Linux installation and the build of ITKv4 along with a collection of exercises to be run during the tutorial. Each one of the sessions will include hands-on exercises on the use of the new ITKv4 functionalities.

Organizer

• Luis Ibanez, Ph.D., Technical Leader, Kitware Inc.