Core 1

Please provide a table of Core 1 algorithms versus investigating NAMIC sites. This table should help us delineate cross-collaboration between NAMIC sites and the types of imaging modalities covered by each site. This table can be placed on the NAMIC Wiki and should be referenced in the year 2 progress report.

The following table describes algorithms that have either been developed, implemented and integrated into Slicer or ITK, or are partially through the process of implementation and integration Algorithm Table

On page 115 of your progress report, you state "The standardization efforts in shape analysis will facilitate the development of tools incorporating methods from different NAMIC sites." Please describe further the standardization efforts and your intentions for the outcomes of Core 1. '

Shape analysis is an example of a population analysis procedure that operates on a group of subjects in the study and produces a single answer that is a model of shape variability or differences in the group. In contrast to many other operations that are subject-specific (for example, segmentation or DTI interpretation), population analysis requires additional software infrastructure to characterize the group(s) of subjects considered in the analysis.

Since the image analysis infrastructure in ITK was originally targeted towards single-subject processing pipelines, we have started development of the fundamental modules and interfaces that will enable description and manipulation of the population data. Our goal was to extend the ITK platform with the population descriptors that allow all NAMIC developers to build population pipelines and to integrate they new algorithms for population analysis into such pipelines.

This project was originally carried out jointly by the UNC and the MIT groups in Core 1; the initial drafts of the infrastructure can be found on the Wiki: Shape Analysis

We have also been in communications with Core 2 developers, and in the spring of 2005 it was decided that Core 2 will incorporate the population descriptors into the basic ITK infrastructure and maintain the code once it has been developed and released as part of the NAMIC Kit. This decision reflects the importance of the population infrastructure and its natural fit with all other building blocks of the ITK code base, previously centered around subject-specific operations. We have made good progress towards completing the first version of such a population support during the Programming Week (June 2005), and expect to have the infrastructure in place by the end of Fall 2005. We will then start implementing pipelines for shape analysis and integrating the specific analysis modules into the NAMIC kit.