Segmentation of brain structures and lesions is a recurrent problem in neuroscience. Software tools that com- bine advanced mathematical methodology with well-designed easy-to-use human interfaces can help neurolo- gists produce more accurate segmentations in less time. ITK-SNAP is a multiplatform open source application that provides an intuitive interface for user-guided segmentation of 3D images using the active contour ap- proach, along with a set of supporting interfaces for manual structure delineation and 3D image navigation. ITK-SNAP has been demonstrated to improve segmentation reliability and efficiency over manual delineation and it is already helping many researchers meet their segmentation needs. ITK-SNAP is available freely through its own Internet site, which supports a growing user community. ? ? The objective of this proposal is to make ITK-SNAP accessible to a much wider audience of neuroscientists by making improvements to its interoperability, usability and documentation. Interoperability enhancements in- clude comprehensive support for the NIFTI and DICOM image formats, with a special emphasis on correct interpretation of the transformations between image space and patient space encoded in NIFTI and DICOM headers. Proposed usability enhancements include the ability to load and process multiple image layers at once, enabling concurrent segmentation of multi-modality imaging data, segmentation of white matter struc- tures in diffusion-weighted MRI, and other segmentation applications for which limited software resources exist today. New documentation will describe several scenarios that combine ITK-SNAP with prominent neuroimaging tools like SPM and AFNI to create task-oriented image analysis pipelines that leverage segmentation. ? ? In addition to making ITK-SNAP more accessible, this proposal aims to integrate it with the NIH Neuroimaging Informatics Tools and Resources Clearinghouse (NITRC). This effort will help expose many more users to the tool and create an environment in which users and developers can interact productively, leading to additional high-impact enhancements to interoperability and usability. This application proposes to improve the interoperability, usability and documentation of ITK-SNAP, a software application that helps neuroanatomists delineate brain structures and lesions in three-dimensional medical im- ages more quickly and more accurately. The proposed effort will make the tool more accessible and available to more biomedical researchers and, subsequently, will have an impact on the cost-efficiency and accuracy of various clinical studies and intervention procedures. ? ? ?
| Yushkevich, Paul A; Avants, Brian B; Pluta, John et al. (2009) A high-resolution computational atlas of the human hippocampus from postmortem magnetic resonance imaging at 9.4 T. Neuroimage 44:385-98 |
