Yushkevich PA, Piven J, Hazlett HC, Smith RG, Ho S, Gee JC, Gerig GUser-guided 3D active contour segmentation of anatomical structures: significantly improved efficiency and reliability. Neuroimage 31:1116-1128

Penn Image Computing and Science Laboratory, Department of Radiology, University of Pennsylvania, PA 19104-6274, USA.
NeuroImage (Impact Factor: 6.36). 08/2006; 31(3):1116-28. DOI: 10.1016/j.neuroimage.2006.01.015
Source: PubMed


Active contour segmentation and its robust implementation using level set methods are well-established theoretical approaches that have been studied thoroughly in the image analysis literature. Despite the existence of these powerful segmentation methods, the needs of clinical research continue to be fulfilled, to a large extent, using slice-by-slice manual tracing. To bridge the gap between methodological advances and clinical routine, we developed an open source application called ITK-SNAP, which is intended to make level set segmentation easily accessible to a wide range of users, including those with little or no mathematical expertise. This paper describes the methods and software engineering philosophy behind this new tool and provides the results of validation experiments performed in the context of an ongoing child autism neuroimaging study. The validation establishes SNAP intrarater and interrater reliability and overlap error statistics for the caudate nucleus and finds that SNAP is a highly reliable and efficient alternative to manual tracing. Analogous results for lateral ventricle segmentation are provided.

Download full-text


Available from: Heather Cody,
1 Follower
124 Reads
    • "The left and right hippocampus were manually segmented on the MPRAGE image, using the interactive software program " ITK-SNAP " version 2.1 [Yushkevich et al., 2006] ( Segmentation was performed using a previously published protocol [Geuze et al., 2005] in which segmentation was performed from posterior to anterior. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Introduction: Cerebral small vessel disease is one of the most important risk factors for dementia, and has been related to hippocampal atrophy, which is among the first observed changes on conventional MRI in patients with dementia. However, these volumetric changes might be preceded by loss of microstructural integrity of the hippocampus for which conventional MRI is not sensitive enough. Therefore, we investigated the relation between the hippocampal diffusion parameters and the risk of incident dementia, using diffusion tensor imaging, independent of hippocampal volume. Methods: The RUNDMC study is a prospective study among 503 elderly with small vessel disease, without dementia, with 5 years follow-up in 2012 (99.6% response-rate). Cox regression analysis was performed to calculate hazard ratios for dementia, of fractional anisotropy and mean diffusivity within the hippocampus, adjusted for demographics, hippocampal volume, and white matter. This was repeated in participants without evident hippocampal volume loss, because in these participants the visible damage might not yet have already started, whereas damage might have started on a microstructural level. Results: 43 participants developed dementia (8.6%), resulting in a 5.5-year cumulative risk of 11.1% (95%CI 7.7-14.6). Higher mean diffusivity was associated with an increased 5-year risk of dementia. In the subgroup of participants with the upper half hippocampal volume, higher hippocampal mean diffusivity, more than doubled the 5-year risk of dementia. Conclusion: This is the first prospective study showing a relation between a higher baseline hippocampal mean diffusivity and the risk of incident dementia in elderly with small vessel disease at 5-year follow-up, independent of hippocampal volume and white matter volume. Hum Brain Mapp, 2015. © 2015 Wiley Periodicals, Inc.
    Human Brain Mapping 10/2015; DOI:10.1002/hbm.23029 · 5.97 Impact Factor
  • Source
    • "Again, the transformation was computed using a symmetric normalized diffeomorphic transformation model. Tract segmentation was achieved by manually drawing a polygon around the DRTT in each histological slice (Yushkevich et al. 2006). Two investigators with a good anatomical knowledge of the DRTT executed the segmentation (J. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Diffusion-weighted imaging (DWI) tractography is a technique with great potential to characterize the in vivo anatomical position and integrity of white matter tracts. Tractography, however, remains an estimation of white matter tracts, and false-positive and false-negative rates are not available. The goal of the present study was to compare postmortem tractography of the dentatorubrothalamic tract (DRTT) by its 3D histological reconstruction, to estimate the reliability of the tractography algorithm in this specific tract. Recent studies have shown that the cerebellum is involved in cognitive, language and emotional functions besides its role in motor control. However, the exact working mechanism of the cerebellum is still to be elucidated. As the DRTT is the main output tract it is of special interest for the neuroscience and clinical community. A postmortem human brain specimen was scanned on a 7T MRI scanner using a diffusion-weighted steady-state free precession sequence. Tractography was performed with PROBTRACKX. The specimen was subsequently serially sectioned and stained for myelin using a modified Heidenhain-Woelke staining. Image registration permitted the 3D reconstruction of the histological sections and comparison with MRI. The spatial concordance between the two modalities was evaluated using ROC analysis and a similarity index (SI). ROC curves showed a high sensitivity and specificity in general. Highest measures were observed in the superior cerebellar peduncle with an SI of 0.72. Less overlap was found in the decussation of the DRTT at the level of the mesencephalon. The study demonstrates high spatial accuracy of postmortem probabilistic tractography of the DRTT when compared to a 3D histological reconstruction. This gives hopeful prospect for studying structure-function correlations in patients with cerebellar disorders using tractography of the DRTT.
    Brain Structure and Function 10/2015; DOI:10.1007/s00429-015-1115-7 · 5.62 Impact Factor
    • "In this work, we use the multi-atlas PatchMatch algorithm for cardiac image segmentation (de Marvao et al., 2014; Shi et al., 2013a). Initially, the left ventricular cavity, the myocardium and the right ventricular cavity were manually segmented in 20 subjects at the ED and ES phase using the software itkSNAP (Yushkevich et al., 2006) by experienced clinicians in accordance with the SCMR recommendations (Schulz-Menger et al., "
    [Show abstract] [Hide abstract]
    ABSTRACT: Atlases encode valuable anatomical and functional information from a population. In this work, a bi-ventricular cardiac atlas was built from a unique data set, which consists of high resolution cardiac MR images of 1000+ normal subjects. Based on the atlas, statistical methods were used to study the variation of cardiac shapes and the distribution of cardiac motion across the spatio-temporal domain. We have shown how statistical parametric mapping (SPM) can be combined with a general linear model to study the impact of gender and age on regional myocardial wall thickness. Finally, we have also investigated the influence of the population size on atlas construction and atlas-based analysis. The high resolution atlas, the statistical models and the SPM method will benefit more studies on cardiac anatomy and function analysis in the future.
    Medical image analysis 09/2015; 26(1):133-145. DOI:10.1016/ · 3.65 Impact Factor
Show more