|
|
| (133 intermediate revisions by 14 users not shown) |
| Line 1: |
Line 1: |
| − | Back to [[2008_Progress_Report]]
| + | This latest draft of this report has now been removed from the wiki and is available [[Media:2008_Namic_Progress_Report.doc|here]] in a MS word document for the final submission. If you have any changes to the last version of text, please send these to Tina. The final version will be posted back here by May 30th. If you really need to look at the last wiki version, please click on the history tab of this page and look at the last one edited on May 22nd. [[User:Tkapur|Tkapur]] 14:07, 23 May 2008 (EDT) |
| − | | |
| − | | |
| − | | |
| − | | |
| − | | |
| − | =Guidelines for preparation=
| |
| − | | |
| − | *[[2008_Progress_Report#Scientific Report Timeline]] - Main point is that May 15 is the date by which all sections below need to be completed. No extensions are possible.
| |
| − | *DBPs - If there is work outside of the roadmap projects that you would like to report, you are welcome to create a separate section for it under "Other".
| |
| − | *The outline for this report is similar to the 2007 report, which is provided here for reference: [[2007_Annual_Scientific_Report]].
| |
| − | *In preparing summaries for each of the 8 topics in this report, please leverage the detailed pages for projects provided here: [[NA-MIC_Internal_Collaborations]].
| |
| − | *Publications will be mined from the SPL publications database. All core PIs need to ensure that all NA-MIC publications are in the publications database by May 15.
| |
| − | | |
| − | =Introduction (Tannenbaum)=
| |
| − | | |
| − | The National Alliance for Medical Imaging Computing (NA-MIC) is now in its fourth year. This Center is comprised of a multi-institutional, interdisciplinary team of computer scientists, software engineers, and medical investigators who have come together to develop and apply computational tools for the analysis and visualization of medical imaging data. A further purpose of the Center is to provide infrastructure and environmental support for the development of computational algorithms and open source technologies, and to oversee the training and dissemination of these tools to the medical research community. The first driving biological projects (DBPs) three years for Center were inspired by schizophrenia research. In the fourth year new DBPs have been added. Three are centered around diseases of the brain: (a) brain lesion analysis in neuropschiatric systemic lupus erythematosus; (b) a study of cortical thickness for autism; and (c) stochastic tractography for VCFS. In an very new direction, we have added DBP on the prostate: brachytherapy needle positioning robot integration.
| |
| − | | |
| − | We briefly summarize the work of NAMIC during the four years of its existence. In the year one of the Center, alliances were forged amongst the cores and constituent groups in order to integrate the efforts of the cores and to define the kinds of tools needed for specific imaging applications. The second year emphasized the identification of the key research thrusts that cut across cores and were driven by the needs and requirements of the DBPs. This led to the formulation of the Center's four main themes: Diffusion Tensor Analysis, Structural Analysis, Functional MRI Analysis, and the integration of newly developed tools into the NA-MIC Tool Kit. The third year of center activity was devoted to the continuation of the collaborative efforts in order to give solutions to the various brain-oriented DBPs.
| |
| − | | |
| − | Year four has seen progress with the work of our new DBPs. As alluded to above these include work on neuropsychiatric disorders such as Systemic Lupus Erythematosis (MIND Institute, University of New Mexico), Velocardiofacial Syndrome (Harvard), and Autism (University of North Carolina, Chapel Hill), as well as the prostate interventional work (Johns Hopkins and Queens Universities). We already have a number of publications as is indicated on our publications page, and software development is continuing as well.
| |
| − | | |
| − | In the next section (Section 3), we summarize this year’s progress on the four roadmap projects listed above: Section 3.1 stochastic tractography for Velocardiofacial Syndrome, Section 3.2 brachytherapy needle positioning for the prostate, Section 3.3 brain lesion analysis in neuropschiatric systemic lupus erythematosus, and Section 3.4 cortical thickness for autism. Next in Section 4, we describe recent work on the four infrastructure topics. These include: Diffusion Image analysis (Section 4.1), Structural analysis (Section 4.2), Functional MRI analysis (Section 4.3), and the NA-MIC Toolkit (Section 4.4). In Section 4.5, we outline some of the other key projects, in Section 4.6 some key highlights including the integration of the EM Segmentor into Slicer, and in Section 4.7 the impact of biocomputing at three different levels: within the center, within the NIH-funded research community, and externally to a national and international community. The final section of this report, Section 4.8, provides a timeline of Center activities.
| |
| − | | |
| − | =Clinical Roadmap Projects=
| |
| − | ==Roadmap Project: Stochastic Tractography for VCFS (Kubicki)==
| |
| − | ===Overview (Kubicki)===
| |
| − | The goal of this project is to create an end-to-end application that would be usefull in evaluating anatomical connectivity between segmented cortical regions of the brain. The ultimate goal of our program is to understand anatomical connectivity similarities and differences between genetically related schizophrenia and velocardio-fatial syndrome. Thus we plan to use the "stochastic tractography" tool for the analysis of abnormalities in integrity, or connectivity, provided by arcuate fasciculus, fiber bundle involved in language processing, in schizophrenia and VCFS.
| |
| − | | |
| − | ===Algorithm Component (Golland)===
| |
| − | Over the last year, algorythm has been tested first on the already available to NAMIC dataset of schizophrenia subjects acquired on 1.5T. This step allowed us to optimize algorythm to our dataset, as well as to develop the pipeline for data analysis that would be then easilly transferable to other datasets, and structures. For this, as well as other applications, we use gray matter labels derived from either automatic or manual segmentation of structural MRI. Thus the first step was to obtain registration between diffusion and anatomical data. Right now, demon's registration alogythm that is part of slicer is being used, but we are also testing b-spline registration, as well as fluid? registration for this purpose. Next step, also accomplished this last year, was to apply the alogythm to new, higher resolution NAMIC dataset, and study smaller white matter connections including cingulum bundle, arcuate fasciculus, uncinate fasciculus and internal capsule. This step was accomplished and data presented at the Santa Fee meeting in October 2007. Algorythm was also additionally tested on the phantom, where differences in coordinate systems were debugged. Upon the completion of testing phase, we started analysis of arcuate fasciculus, language related fiber bundle, in new 3T, high resolution dataset. Additional step turned out to be required in preprocessing, in order to prevent algorithm from traveling through the ventricles, where diffusivity is high, thus white matter sesgmentation was performed using free surfer, and co-registered with DTI. Our current work focuses on better registration alogythms, as well as the way to parametrize tracts, in order to obtain FA measurements along the tracts.
| |
| − | | |
| − | ===Engineering Component (Davis)===
| |
| − | ===Clinical Component (Kubicki)===
| |
| − | Over the last year, algorythm has been tested first on the already available to NAMIC dataset of schizophrenia subjects acquired on 1.5T. Anterior Limb of the internal capsule, large structure connecting thalamus with frontal lobe, have been extracvted, and analyzed in group of 20 schizphrenics, and 20 control subjects, and results showing group differences in Fractional Anisotropy presented at the ACNP symposium in December 2007. Next, stochastic tractography was tested, and optimized for new, high resolution DTI dataset acquired on 3T GE magnet. Upon the completion of testing phase, we started analysis of arcuate fasciculus, language related fiber bundle, in 20 controls and 20 chronic schizphrenics. Whole brain as well as white matter segmentations, using freesurfer, as well as automatic extraction of regions interconnected by Arcuate Fasciculus (Inferior frontal and Supperior Temporal Gyri), as well as another ROI that would guide the tract ("waypoint" ROI) have been generated for all subjects. Finally, paths of interest were generated, and averaged FA extracted for each tract. Preliminary data based on 7 patients and 12 controls were presented at the AHM in January 2008, study is currently under way.
| |
| − | | |
| − | ===Additional Information===
| |
| − | Additional Information for this project is available [http://wiki.na-mic.org/Wiki/index.php/DBP2:Harvard:Brain_Segmentation_Roadmap here on the NA-MIC wiki].
| |
| − | ==Roadmap Project: Brachytherapy Needle Positioning Robot Integration (Fichtinger)==
| |
| − | ===Overview (Fichtinger)===
| |
| − | ===Algorithm Component (Tannenbaum)===
| |
| − | We have worked on both the segmentation and the registration of the prostate from MRI and ultrasound data. We explain each of the steps now.
| |
| − | | |
| − | ====Prostate Segmentation====
| |
| − | | |
| − | We first must extract the prostate. We have considered three possible methods: a combination of a combination of Cellular Automata(CA also known as Grow Cut) with Geometric Active Contour(GAC) methods; employing an ellipsoid to match the prostate in 3D image; shape based approach using spherical wavelets. More details are given below and images and further details may be found at [[Projects:ProstateSegmentation|GaTech Algorithm Prostate Segmentation]].
| |
| − | | |
| − | 1. A cellular automata algorithm is used to give an initial segmentation. It begins with a rough manual initialization and then iteratively classifies all pixels into object and bacground until convergence. It effectively overcomes the problems of weak boundaries and inhomogeneity within the object or background. This in turn is fed into Geometric Active Contour for finer tuning. We are initially using the edge-based minimal surface pproach (the generalization of the standard Geodesic Active Contour model) which seems to give very reasonable results. Both steps of the algorithm algorithm are implemented in 3D. A ITK-Cellular Automata filter, dealing with N-D data, has already been completed and submitted to the NA-MIC SandBox.
| |
| − | | |
| − | 2. Spherical wavelets have proven to be a very natural way of representing 3D shapes which are compact and simply connected (topological spheres). We developed a segmentation framework using this 3D wavelet representation and multiscale prior. The parameters of our model are the learned shape parameters based on the spherical wavelet coefficients}, as well as pose parameters that accommodate for shape variability due to a similarity transformation (rotation, scale, translation) which is not explicitly modeled with the shape parameters. The transformed surface based on the pose parameters. We used a region-based energy to drive the evolution of the parametric deformable surface for segmentation. Our segmentation algorithm deforms an initial surface according to the gradient flow that minimizes the energy functional in terms of the pose and shape parameters. Additionally, the optimization method can be applied in a coarse to fine manner. Spherical wavelets and conformal mappings are
| |
| − | already part of the NA-MIC SandBox.
| |
| − | | |
| − | 3. The third method is very closely related to the second. It is based on the observation that the prostate may be roughly modelled as an ellipsoid. One can then employing this ellipsoid model coupled with a local/global segmentation energy approach which we have developed this year, as the basis of a segmentation procedure. Because of the local/global nature of the functional and the implicit introduction of scale this methodology may be very useful for MRI prostate data.
| |
| − | | |
| − | ====Prostate Registration====
| |
| − | | |
| − | The registration and segmentation elements of our algorithm are difficult to separate. Thus for the 3D shape-driven segmentation part, the shapes must first be aligned through a conformal and area-correction alignment process. The prostate presents a number of difficulties for traditional approaches since there are no easily discernable landmarks. On the other hand, we observed that the surface of the prostate is almost half convex and half concave. The concave region may be captured and used to register the shapes, thus we register the whole shape by registering a certain region on it. Such concave region is characterized by its negative mean curvature. We treat the mean curvature as a scalar field defined on the surface, and we have extended the Chan-Vese method (in which one wants to separate the means with respect to the regions defined by the interior and exterior of the evolving active contour) to the case at hand on the prostate surface. The method is implemented in C++ and it successfully extracts the concave surface region. This method could also be used to exact regions on surface according to any feature charactered by a scalar field defined on the surface.
| |
| − | | |
| − | In order incorporate the extracted region as landmarks into the registration process, instead of matching two binary images directly, we transform the binary images into a form to highlight the boundary region. This is done by applying a Gauss function on the (narrow band) of the signed distance function of the binary image. The transformed image enjoys the advantages of both the parametric and implicit representations of shapes. Namely it has compact description, as the parametric representation does, and as in the implicit representation it avoids the correspondence problem. Moreover we incorporate the extracted concave regions into such images for registration which leads to a better result.
| |
| − | | |
| − | Finally, in the past year we have developed a particle filtering approach for the general problem of registering two point sets that differ by a rigid body transformation which may be very useful for this project. Typically, registration algorithms compute the transformation parameters by maximizing a metric given an estimate of the correspondence between points across the two sets of interest. This can be viewed as a posterior estimation problem, in which the corresponding distribution can naturally be estimated using a particle filter. We treat motion as a local variation in pose parameters obtained from running several iterations of the standard Iterative Closest Point (ICP) algorithm. Employing this idea, we introduce stochastic motion dynamics to widen the narrow band of convergence often found in local optimizer functions used to tackle the registration task. In contrast with other techniques, this approach requires no annealing schedule, which results in a reduction in computational complexity as well as maintains the temporal coherency of the state (no loss of information). Also, unlike most alternative approaches for point set registration, we make no geometric assumptions on the two data sets.
| |
| − | | |
| − | ===Engineering Component (Hayes)===
| |
| − | ===Clinical Component (Fichtinger)===
| |
| − | ===Additional Information===
| |
| − | Additional Information for this project is available [http://wiki.na-mic.org/Wiki/index.php/DBP2:JHU:Roadmap here on the NA-MIC wiki].
| |
| − | ==Roadmap Project: Brain Lesion Analysis in Neuropsychiatric Systemic Lupus Erythematosus (Bockholt)==
| |
| − | ===Overview (Bockholt)===
| |
| − | ===Algorithm Component (Whitaker)===
| |
| − | ===Engineering Component (Pieper)===
| |
| − | ===Clinical Component (Bockholt)===
| |
| − | ===Additional Information===
| |
| − | Additional Information for this project is available [http://wiki.na-mic.org/Wiki/index.php/DBP2:MIND:Roadmap here on the NA-MIC wiki].
| |
| − | ==Roadmap Project: Cortical Thickness for Autism(Hazlett)==
| |
| − | ===Overview (Hazlett)===
| |
| − | | |
| − | We would like to create an end-to-end application within Slicer3 allowing individual and group analysis of regional and local cortical thickness. Such a workflow applied to our study comparing healthy control to autistic children in the young brain (2-4 years old) is the main goal of the UNC DBP.
| |
| − | | |
| − | ===Algorithm Component (Styner)===
| |
| − | | |
| − | The basic steps necessary for the cortical thickness application entail first tissue segmentation in order to separate white and gray matter regions, second cortical thickness measurement, thirdly cortical correspondence to compare measurements across subjects and finally a statistical analysis to locally compute group differences. | |
| − | Tissue segmentation: We have successfully adapted the UNC segmentation tool called itkEMS to Slicer, which we have for segmentations of the young brain. We also created a young brain atlas for the current Slicer3 EM Segment module. Tests have been successful and a comparative study to itkEMS has shown that further parameter optimization is needed to reach the same quality.
| |
| − | | |
| − | ====Cortical thickness measurement====
| |
| − | The UNC algorithm for the measurement of local cortical thickness given a labeling of white matter and gray matter has been developed into a Slicer3 external module. This module lends itself well for regional analysis of cortical thickness, but less so for local analysis due to its non-symmetric and sparse measurements. Ongoing development is focusing on a symmetric, Laplacian based cortical thickness suitable for local analysis.
| |
| − | | |
| − | ====Cortical correspondence (regional)====
| |
| − | | |
| − | For regional correspondence, an existing lobar parcellation atlas is deformably registered using a b-spline registration tool. First tests have been very promising and the release of the corresponding Slicer 3 registration module is schedule to be finished within the next month and thus the regional analysis workflow will be available at that time.
| |
| − | | |
| − | ====Cortical correspondence (local)====
| |
| − | Local cortical correspondence requires a two-step process of white/gray surface inflation followed by group-wise correspondence computation. White matter surface extraction and inflation is currently achieved with an external tool and developing a Slicer 3 based solution is a goal in the next year. The group-wise correspondence step has been fully solved, and a Slicer 3 module is already available. Evaluation on real data has shown that our method outperforms the currently widely employed Freesurfer framework.
| |
| − | | |
| − | ====Statistical analysis/Hypothesis testing====
| |
| − | Regional analysis can be done with standard statistical tools such as MANOVA as there are a limited, relatively small number of regions. Local analysis on the other hand needs local non-parametric testing, multiple-comparison correction, and correlative analysis that is not routinely available. We are currently extending the current Slicer 3 module designed for statistical shape analysis to be used for this purpose incorporating a local applied General Linear Module and MANCOVA based testing framework.
| |
| − | | |
| − | ===Engineering Component (Miller, Vachet)===
| |
| − | ===Clinical Component (Hazlett)===
| |
| − | ===Additional Information===
| |
| − | Additional Information for this project is available [http://wiki.na-mic.org/Wiki/index.php/DBP2:UNC:Cortical_Thickness_Roadmap here on the NA-MIC wiki].
| |
| − | | |
| − | =Four Infrastructure Topics=
| |
| − | ==Diffusion Image Analysis (Gerig)==
| |
| − | ===Progress===
| |
| − | ===Key Investigators===
| |
| − | ===Additional Information===
| |
| − | Additional Information for this topic is available [http://wiki.na-mic.org/Wiki/index.php/NA-MIC_Internal_Collaborations:DiffusionImageAnalysis here on the NA-MIC wiki].
| |
| − | ==Structural Analysis(Tannenbaum)==
| |
| − | ===Progress===
| |
| − | Under Structural Analysis, the main topics of research for NAMIC are structural segmentation, registration techniques and shape analysis. These topics are correlated and research in one often finds application in another. For example, shape analysis can yield useful priors for segmentation, or segmentation and registration can provide structural correspondences for use in shape analysis and so on.
| |
| − | | |
| − | An overview of selected progress highlights under these broad topics follows.
| |
| − | | |
| − | Structural Segmentation
| |
| − | | |
| − | * Directional Based Segmentation
| |
| − | We have proposed a directional segmentation framework for Direction-weighted Magnetic Resonance imagery by augmenting the Geodesic Active Contour framework with directional information. The classical scalar conformal factor is replaced by a factor that incorporates directionality. We mathematically showed that the optimization problem is well-defined when the factor is a Finsler metric. The calculus of variations or dynamic programming may be used to find the optimal curves. This past year we have applied this methodology in extracting the anchor tract (or centerline) of neural fiber bundles. Further we have applied this in conjunction with the Bayes’ rule into volumetric segmentation for extracting the entire fiber bundles. We have also proposed a novel shape prior in the volumetric segmentation to extract tubular fiber bundles.
| |
| − | | |
| − | * Stochastic Segmentation
| |
| − | | |
| − | We have continued work this year on developing new stochastic methods for implementing curvature-driven flows for medical tasks like segmentation. We can now generalize our results to an arbitrary Riemannian surface which includes the geodesic active contours as a special case. We are also implementing the directional flows based on the anisotropic conformal factor described above using this stochastic methodology. Our stochastic snakes’ models are based on the theory of interacting particle systems. This brings together the theories of curve evolution and hydrodynamic limits, and as such impacts our growing use of joint methods from probability and partial differential in image processing and computer vision. We now have working code written in C++ for the two dimensional case and have worked out the stochastic model of the general geodesic active contour model.
| |
| − | | |
| − | * Statistical PDE Methods for Segmentation
| |
| − | | |
| − | Our objective is to add various statistical measures into our PDE flows for medical imaging. This will allow the incorporation of global image information into the locally defined PDE framework. This year, we developed flows which can separate the distributions inside and outside the evolving contour, and we have also been including shape information in the flows. We have completed a statistically based flow for segmentation using fast marching, and the code has been integrated into Slicer.
| |
| − | | |
| − | * Atlas Renormalization for Improved Brain MR Image Segmentation
| |
| − | | |
| − | Atlas-based approaches can automatically identify detailed brain structures from 3-D magnetic resonance (MR) brain images. However, the accuracy often degrades when processing data acquired on a different scanner platform or pulse sequence than the data used for the atlas training. In this project, we work to improve the performance of an atlas-based whole brain segmentation method by introducing an intensity renormalization procedure that automatically adjusts the prior atlas intensity model to new input data. Validation using manually labeled test datasets shows that the new procedure improves segmentation accuracy (as measured by the Dice coefficient) by 10% or more for several structures including hippocampus, amygdala, caudate, and pallidum. The results verify that this new procedure reduces the sensitivity of the whole brain segmentation method to changes in scanner platforms and improves its accuracy and robustness, which can thus facilitate multicenter or multisite neuroanatomical imaging studies.
| |
| − | | |
| − | *Multiscale Shape Segmentation Techniques
| |
| − | | |
| − | The goal of this project is to represent multiscale variations in a shape population in order to drive the segmentation of deep brain structures, such as the caudate nucleus or the hippocampus. Our technique defines a multiscale parametric model of surfaces belonging to the same population using a compact set of spherical wavelets targeted to that population. We derived a parametric active surface evolution using the multiscale prior coefficients as parameters for our optimization procedure to naturally include the prior for segmentation. Additionally, the optimization method can be applied in a coarse-to-fine manner. We applied our algorithm to the caudate nucleus, a brain structure of interest in the study of schizophrenia. Our validation shows that our algorithm is computationally efficient and outperforms the Active Shape Model (ASM) algorithm, by capturing finer shape details.
| |
| − | | |
| − | Registration
| |
| − | | |
| − | * Optimal Mass Transport Registration
| |
| − | The aim of this project is to provide a computationally efficient non-rigid/elastic image registration algorithm based on the Optimal Mass Transport theory. We use the Monge-Kantorovich formulation of the Optimal Mass Transport problem and implement the gradient flow PDE approach using multi-resolution and multi-grid techniques to speed up the convergence. We also leverage the computational power of general purpose graphics processing units available on standard desktop computing machines to exploit the inherent parallelism in our algorithm. We have implemented 2D and 3D multi-resolution registration using Optimal Mass Transport and are currently working on the registration of 3D datasets.
| |
| − | | |
| − | * Diffusion Tensor Image Processing Tools
| |
| − |
| |
| − | We aim to provide methods for computing geodesics and distances between diffusion tensors. One goal is to provide hypothesis testing for differences between groups. This will involve interpolation techniques for diffusion tensors as weighted averages in the metric framework. We will also provide filtering and eddy current correction. This year, we developed a Slicer module for DT-MRI Rician noise removal, developed prototypes of DTI geometry and statistical packages, and began work on a general method for hypothesis testing between diffusion tensor groups.
| |
| − | | |
| − | * Point Set Rigid Registration
| |
| − | | |
| − | We propose a particle filtering scheme for the registration of 2D and 3D point set undergoing a rigid body transformation where we incorporate stochastic dynamics to model the uncertainty of the registration process. Typically, registration algorithms compute the transformations parameters by maximizing a metric given an estimate of the correspondence between points across the two sets of interest. This can be viewed as a posterior estimation problem, in which the corresponding distribution can naturally be estimated using a particle filter. In this work, we treat motion as a local variation in the pose parameters obtained from running a few iterations of the standard Iterative Closest Point (ICP) algorithm. Employing this idea, we introduce stochastic motion dynamics to widen the narrow band of convergence as well as provide a dynamical model of uncertainty. In contrast with other techniques, our approach requires no annealing schedule, which results in a reduction in computational complexity as well as maintains the temporal coherency of the state (no loss of information). Also, unlike most alternative approaches for point set registration, we make no geometric assumptions on the two data sets.
| |
| − | | |
| − | * Cortical Correspondence using Particle System
| |
| − | | |
| − | In this project, we want to compute cortical correspondence on populations, using various features such as cortical structure, DTI connectivity, vascular structure, and functional data (fMRI). This presents a challenge because of the highly convoluted surface of the cortex, as well as because of the different properties of the data features we want to incorporate together. We would like to use a particle based entropy minimizing system for the correspondence computation, in a population-based manner. This is advantageous because it does not require a spherical parameterization of the surface, and does not require the surface to be of spherical topology. It would also eventually enable correspondence computation on the subcortical structures and on the cortical surface using the same framework. To circumvent the disadvantage that particles are assumed to lie on local tangent planes, we plan to first ‘inflate’ the cortex surface. Currently, we are at testing stage using structural data, namely, point locations and sulcal depth (as computed by FreeSurfer).
| |
| − | | |
| − | * Multimodal Atlas
| |
| − | | |
| − | In this work, we propose and investigate an algorithm that jointly co-registers a collection of images while computing multiple templates. The algorithm, called iCluster for Image Clustering, is based on the following idea: given the templates, the co-registration problem becomes simple, reducing to a number of pairwise registration instances. On the other hand, given a collection of images that have been co-registered, an off-the shelf clustering or averaging algorithm can be used to compute the templates. The algorithm assumed a fixed and known number of template images. We formulate the problem as a maximum likelihood solution and employ a Generalized Maximum Likelihood algorithm to solve it. In the E-step, we compute membership probabilities. In the M-step, we update the template images as weighted averages of the images, where weights are the memberships and the template priors are updated, and then perform a collection of independent pairwise registration instances. The algorithm is currently implemented in the Insight ToolKit (ITK) and we next plan to integrate it into Slicer.
| |
| − | | |
| − | * Groupwise Registration
| |
| − | | |
| − | We aim at providing efficient groupwise registration algorithms for population analysis of anatomical structures. Here we extend a previously demonstrated entropy based groupwise registration method to include a free-form deformation model based on B-splines. We provide an efficient implementation using stochastic gradient descents in a multi-resolution setting. We demonstrate the method in application to a set of 50 MRI brain scans and compare the results to a pairwise approach using segmentation labels to evaluate the quality of alignment. Our results indicate that increasing the complexity of the deformation model improves registration accuracy significantly, especially at cortical regions.
| |
| − | | |
| − | Shape Analysis
| |
| − | | |
| − | * Shape Analysis Framework Using SPHARM-PDM
| |
| − | | |
| − | The UNC shape analysis is based on an analysis framework of objects with spherical topology, described by sampled spherical harmonics SPHARM-PDM. The input of the proposed shape analysis is a set of binary segmentations of a single brain structure, such as the hippocampus or caudate. Group tests can be visualized by P-values and by mean difference magnitude and vector maps, as well as maps of the group covariance information. The implementation has reached a stable framework and has been disseminated to several collaborating labs within NAMIC (BWH, Georgia Tech, Utah). The current development focuses on integrating the current command line tools into the Slicer (v3) via the Slicer execution model. The whole shape analysis pipeline is encapsulated and accessible to the trained clinical collaborator. The current toolset distribution (via NeuroLib) now also contains open data for other researchers to evaluate their shape analysis enhancements.
| |
| − | | |
| − | * Multiscale Shape Analysis
| |
| − | | |
| − | We present a novel method of statistical surface-based morphometry based on the use of non-parametric permutation tests and a spherical wavelet (SWC) shape representation. As an application, we analyze two brain structures, the caudate nucleus and the hippocampus. We show that the results nicely complement the results obtained with shape analysis using a sampled point representation (SPHARM-PDM). We used the UNC pipeline to pre-process the images, and for each triangulated SPHARM-PDM surface, a spherical wavelet description is computed. We then use the UNC statistical toolbox to analyze differences between two groups of surfaces described by the features of choice that is the 3D spherical wavelet coefficients. This year, we conducted statistical shape analysis of the two brain structures and compared the results obtained to shape analysis using a SPHARM-PDM representation.
| |
| − | | |
| − | * Population Analysis of Anatomical Variability
| |
| − | | |
| − | In contrast to shape-based segmentation that utilizes a statistical model of the shape variability in one population (typically based on Principal Component Analysis), we are interested in identifying and characterizing differences between two sets of shape examples. We use the discriminative framework to characterize the differences in shape by training a classifier function and studying its sensitivity to small perturbations in the input data. An additional benefit is that the resulting classifier function can be used to label new examples into one of the two populations, e.g., for early detection in population screening or prediction in longitudinal studies. We have implemented stand alone code for training a classifier, jackknifing and permutation testing, and are currently porting the software into ITK. We have also started exploring alternative, surface-based descriptors which are promising in improving our ability to detect and characterize subtle differences in the shape of anatomical structures due to diseases such as schizophrenia.
| |
| − | | |
| − | * Shape Analysis with Overcomplete Wavelets
| |
| − | | |
| − | In this work, we extend the Euclidean wavelets to the sphere. The resulting over-complete spherical wavelets are invariant to the rotation of the spherical image parameterization. We apply the over-complete spherical wavelet to cortical folding development and show significantly consistent results as well as improved sensitivity compared with the previously used bi-orthogonal spherical wavelet. In particular, we are able to detect developmental asymmetry in the left and right hemispheres.
| |
| − | | |
| − | *Shape based Segmentation and Registration
| |
| − | | |
| − | When there is little or no contrast along boundaries of different regions, standard image segmentation algorithms perform poorly and segmentation is done manually using prior knowledge of shape and relative location of underlying structures. We have proposed an automated approach guided by covariant shape deformations of neighboring structures, which is an additional source of prior knowledge. Captured by a shape atlas, these deformations are transformed into a statistical model using the logistic function. The mapping between atlas and image space, structure boundaries, anatomical labels, and image inhomogeneities are estimated simultaneously within an expectation-maximization formulation of the maximum a posteriori Probability (MAP) estimation problem. These results are then fed into an Active Mean Field approach, which views the results as priors to a Mean Field approximation with a curve length prior. Our method filters out the noise as compared to thresholding using initial likelihoods, and it captures multiple structures as in the brain (where both major brain compartments and subcortical structures are obtained) because it naturally evolves families of curves. The algorithm is currently implemented in 3D Slicer Version 2.6 and a beta version is available in 3D Slicer Version 3.
| |
| − | | |
| − | *Spherical Wavelets
| |
| − | | |
| − | In this project, we apply a spherical wavelet transformation to extract shape features of cortical surfaces reconstructed from magnetic resonance images (MRI) of a set of subjects. The spherical wavelet transformation can characterize the underlying functions in a local fashion in both space and frequency, in contrast to spherical harmonics that have a global basis set. We perform principal component analysis (PCA) on these wavelet shape features to study patterns of shape variation within normal population from coarse to fine resolution. In addition, we study the development of cortical folding in newborns using the Gompertz model in the wavelet domain, allowing us to characterize the order of development of large-scale and finer folding patterns independently. We develop an efficient method to estimate the regularized Gompertz model based on the Broyden–Fletcher–Goldfarb–Shannon (BFGS) approximation. Promising results are presented using both PCA and the folding development model in the wavelet domain. The cortical folding development model provides quantitative anatomical information regarding macroscopic cortical folding development and may be of potential use as a biomarker for early diagnosis of neurological deficits in newborns.
| |
| − | | |
| − | ===Key Investigators===
| |
| − | * MIT: Polina Golland, Kilian Pohl, Sandy Wells, Eric Grimson, Mert R. Sabuncu
| |
| − | * UNC: Martin Styner, Ipek Oguz, Xavier Barbero
| |
| − | * Utah: Ross Whitaker, Guido Gerig, Suyash Awate, Tolga Tasdizen, Tom Fletcher, Joshua Cates, Miriah Meyer
| |
| − | * GaTech: Allen Tannenbaum, John Melonakos, Vandana Mohan, Tauseef ur Rehman, Shawn Lankton, Samuel Dambreville, Yi Gao, Romeil Sandhu, Xavier Le Faucheur, James Malcolm
| |
| − | * Isomics: Steve Pieper
| |
| − | * GE: Bill Lorensen, Jim Miller
| |
| − | * Kitware: Luis Ibanez, Karthik Krishnan
| |
| − | * UCLA: Arthur Toga, Michael J. Pan, Jagadeeswaran Rajendiran
| |
| − | * BWH: Sylvain Bouix, Motoaki Nakamura, Min-Seong Koo, Martha Shenton, Marc Niethammer, Jim Levitt, Yogesh Rathi, Marek Kubicki, Steven Haker
| |
| − | | |
| − | ===Additional Information===
| |
| − | Additional Information for this topic is available [http://wiki.na-mic.org/Wiki/index.php/NA-MIC_Internal_Collaborations:StructuralImageAnalysis here on the NA-MIC wiki].
| |
| − | ==fMRI Analysis (Golland)==
| |
| − | ===Progress===
| |
| − | ===Key Investigators===
| |
| − | ===Additional Information===
| |
| − | Additional Information for this topic is available [http://wiki.na-mic.org/Wiki/index.php/NA-MIC_Internal_Collaborations:fMRIAnalysis here on the NA-MIC wiki].
| |
| − | ==NA-MIC Kit Theme (Schroeder)==
| |
| − | ===Progress===
| |
| − | ===Key Investigators===
| |
| − | * Kitware - Will Schroeder (Core 2 PI), Sebastien Barre, Luis Ibanez, Bill Hoffman
| |
| − | * GE - Jim Miller
| |
| − | * Isomics - Steve Pieper
| |
| − | | |
| − | ===Additional Information===
| |
| − | Additional Information for this topic is available [http://wiki.na-mic.org/Wiki/index.php/NA-MIC-Kit here on the NA-MIC wiki].
| |
| − | ==Other Projects==
| |
| − | Any Project(s) not covered by the 8 sections above
| |
| − | | |
| − | ==Highlights(Schroeder)==
| |
| − | ===EM Segmenter or TBD===
| |
| − | ===DTI progress or TBD===
| |
| − | ===Outreach (Gollub)===
| |
| − | | |
| − | ==Impact and Value to Biocomputing (Miller)==
| |
| − | ===Impact within the Center===
| |
| − | ===Impact within NIH Funded Research===
| |
| − | ===National and International Impact===
| |
| − | ==NA-MIC Timeline (Whitaker)==
| |
| − | | |
| − | ==Appendix A Publications (Kapur)==
| |
| − | These will be mined from the SPL publications database. All core PIs need to ensure that all NA-MIC publications are in the publications database by May 15.
| |
| − | | |
| − | ==Appendix B EAB Report and Response (Kapur)==
| |
| − | ===EAB Report===
| |
| − | ===Response to EAB Report===
| |
