MICCAI 2010

MICCAI 2010 Tutorial Proposal (draft under CONSTRUCTION)

MICCAI main page: http://www.miccai2010.org/

From MICCAI Algorithms to Clinical Translational Tools: The NA-MIC Platform

Outline: MICCAI offers a window on the technological breakthroughs in medical image computing, computer assisted interventions and surgical robotics that will allow for revolutionary progress in clinical care. A critical step to maximize the long-term positive impact of MICCAI algorithms is to foster their deployment, validation and concrete use by the clinical research community for whom they were developed. The objective of the tutorial is to introduce the challenges inherent in delivering advanced medical imaging technologies (image processing, analysis and interventional capabilities) to end users for clinical translational research and to present solutions available in the National Alliance for Medical Image Computing (NA-MIC) open-source software toolkit. The day begins with an overview of the software components of the NA-MIC kit, followed by multiple example applications in current use for clinical research, with a final optional hands-on tutorial session that guides participants through the process of building, testing and deploying an exemplar algorithm using the NA-MIC open-source software toolkit. By the end of the day course participants will know how to use the NA-MIC kit to facilitate greater use of their own algorithms by clinical end users.

Tutorial Faculty

• Ron Kikinis, M.D.
• Guido Gerig, Ph.D.
• Killian Pohl, Ph.D.
• Steve Pieper, Ph.D.
• Noby Hata, Ph.D.
• Dinggang Shen, Ph.D.
• Stephan Aylward, Ph.D.
• Sonia Pujol, Ph.D.
• Xiaodong Tao, Ph.D.
• Martin Styner, Ph.D.
• Sylvain Jaume, Ph.D.?

Workshops academic objectives and justification of relevance to MICCAI

The emergence of increasingly sophisticated mathematical models, image analysis and visualization tools that have followed the rapid development of new medical imaging technologies has led to a better understanding of organ functions in human health and disease. For the past five years, the National Alliance for Medical Image Computing (NA-MIC), one of the seven National Centers for Biomedical Computing (NCBC), part of the NIH Roadmap for medical research, has focused its efforts on the conversion of scientific advances from the biomedical imaging community into an open-source toolkit, so as to improve the availability and deployment of advanced software tools on a national scale.

The NA-MIC toolkit is a set of open-source software packages for medical image computing, which includes 3D Slicer, VTK, ITK, CMake, CTest and BatchMake. 3D Slicer is an end-user application for image analysis and an open-source environment for software development. This deployment platform has been envisioned and designed as a modular software tool that is both easy to use for clinical researchers and easy to extend for programmers.

This tutorial educates participants on the most common obstacles to translating their novel image analysis and computer assisted interventions into clinical research practice and provides open source, freely available solutions to these problems. Multiple examples of clinical research projects that have been recently completed or that are underway are provided to illustrate these issues and solutions. The final optional session is a Hands-On tutorial on how to integrate the NA-MIC toolkit with a third party executable program.

The course uses an exemplar program and pre-computed training MRI datasets. The workshop is intended for scientists and engineers of the medical image analysis community. Pre-requisites include C++ and ITK programming experience.

We believe that building upon the latest open-source analysis tools will help scientists and engineers achieve new technological breakthroughs that will benefit the MICCAI community.

Tutorial format

The tutorial consists of lectures and an optional instructor-led hands-on session with the participants programming on their own laptop computers. A technical training staff will be providing one-to-one assistance as needed.

Agenda:

• 9:00 am - 9:15 am Overview of the NA-MIC Open-Source Software Toolkit (Kikinis)
• 9:15 am - 11:30 am: Lectures
  
• 9:15 am - 9:45 am: Execution Model for Command Line Programs, batch processing (Pieper/Kikinis/Pujol)
  
• 9:45 am - 10:15 am: ‘Hardening’ your Software with NA-MIC Kit, regression tests, dashboard, etc (Aylward)
  
• 10:15 am - 10:30 am: Coffee break
• 10:30 am - 11:00 am: Building bridges among complementary platforms (e.g. Freesurfer, BRAINS2, HAMMER) (Pieper/Pujol)
  
• 11:00 am - 11:30 am: How to participate in NA-MIC – collaboration grants, user and developer groups (Kikinis)
  
• 11:30 am - 12:30 PM: Lunch on your own
• 12:30 pm- 3:45 pm: Exemplar Clinical Research Applications
  
• 12:30 pm - 1:00 pm: Registration (Dinggang Shen)
  
• 1:00 pm - 1:30 pm: Cortical Thickness measurements (Styner)
  
• 1:30 pm - 2:00 pm: EM Segmentation (Gerig/Pohl)
  
• 2:00 pm- 2:30 pm: Diffusion Tensor Imaging (Gerig/Pujol/Styner)
  
• 2:30 pm- 3:00 pm: IGT (Hata)- if no parallel IGT session is submitted
  
• 3:00 pm- 3:15 pm: Coffee break
• 3:15 pm- 3:45 pm Other examples from Community (Ron to send out email to solicit)
• 3:45 pm- 5:00 pm: Optional Hands-on Session: Integrating your code with the NA-MIC Kit (Hello World tutorial extended to include Python and Extensions model) (Pujol)
• 5:00 pm- 6:00 pm: Open Discussion and 1:1 Consultations for individual projects

Anticipated number of participants with reasons (or number from previous workshops)

The NA-MIC consortium has made a strong commitment to education and technology transfer by developing a compendium of training materials on the image analysis tools of the NA-MIC kit. Over the past five years, we have delivered 40 hands-on workshops to more than 900 participants, with an average number of 35 participants per workshop. Given the rapid enrollment in our previous workshops and a close correspondence between the goals of our educational materials and the interests of MICCAI attendees, we anticipate that the MICCAI workshop will gather between 30 and 40 participants.

Space requirements and special equipment or technical support needed

The equipment required includes a classroom setting of tables to seat 30-40 attendees with room for them to place their laptop computers, two projectors, and an Internet connection for the instructor.

Ideas for a MICCAI tutorial

• Goal: Create a program for a MICCAI Turorial to explain

how the MICCAI Community would benefit from the engineering effort and algorithmic contributions accumulated in the NA-MIC Kit for automated image processing. I anticipate that the potential attendees would come from different fields:

- neuroscience (need for accurate brain segmentation and registration at the cost of run jobs overnight),
- image-based clinical trials (need for automated image processing on large cohort of patients or animals), and
- microscopy imaging (need to split the data into blocks that fit into memory and apply the same processing to each block)

We could walk the attendee through the NA-MIC Kit toolbox and explain how to port the image processing task they need (if not already in the NA-MIC Kit) into a Slicer module. For this second aim, I would first demonstrate how to prototype the module using the XML-based user interface. Then I would set up automated builds and testing to process data overnight on the participant's laptop. Finally I could explain how to port the same setup to a server and schedule large scale automated processing.

The message to the MICCAI Committee would be that our tutorial will allow attendees to make their algorithms available to a larger audience (collaborators and non-collaborators) and to develop new algorithms for automated applications. Increasing the visibility of existing MICCAI publications and extending the scope for future MICCAI conferences would be two convincing arguments.