Publications by Year: 2009


Gholami B, Haddad WM, Tannenbaum AR. Agitation and pain assessment using digital imaging. Conf Proc IEEE Eng Med Biol Soc. 2009;2009:2176–9.
Pain assessment in patients who are unable to verbally communicate with medical staff is a challenging problem in patient critical care. The fundamental limitations in sedation and pain assessment in the intensive care unit (ICU) stem from subjective assessment criteria, rather than quantifiable, measurable data for ICU sedation and analgesia. This often results in poor quality and inconsistent treatment of patient agitation and pain from nurse to nurse. Recent advancements in pattern recognition techniques using a relevance vector machine algorithm can assist medical staff in assessing sedation and pain by constantly monitoring the patient and providing the clinician with quantifiable data for ICU sedation. In this paper, we show that the pain intensity assessment given by a computer classifier has a strong correlation with the pain intensity assessed by expert and non-expert human examiners.
Banks DC, Beason KM. Decoupling illumination from isosurface generation using 4D light transport. IEEE Trans Vis Comput Graph. 2009;15(6):1595–602.
One way to provide global illumination for the scientist who performs an interactive sweep through a 3D scalar dataset is to pre-compute global illumination, resample the radiance onto a 3D grid, then use it as a 3D texture. The basic approach of repeatedly extracting isosurfaces, illuminating them, and then building a 3D illumination grid suffers from the non-uniform sampling that arises from coupling the sampling of radiance with the sampling of isosurfaces. We demonstrate how the illumination step can be decoupled from the isosurface extraction step by illuminating the entire 3D scalar function as a 3-manifold in 4-dimensional space. By reformulating light transport in a higher dimension, one can sample a 3D volume without requiring the radiance samples to aggregate along individual isosurfaces in the pre-computed illumination grid.
Pohl KM, Sabuncu MR. A unified framework for MR based disease classification. Inf Process Med Imaging. 2009;21:300–13.
In this paper, we employ an anatomical parameterization of spatial warps to reveal structural differences between medical images of healthy control subjects and disease patients. The warps are represented as structure-specific 9-parameter affine transformations, which constitute a global, non-rigid mapping between the atlas and image coordinates. Our method estimates the structure-specific transformation parameters directly from medical scans by minimizing a Kullback-Leibler divergence measure. The resulting parameters are then input to a linear Support Vector Machine classifier, which assigns individual scans to a specific clinical group. The classifier also enables us to interpret the anatomical differences between groups, as we can visualize the discriminative warp that best differentiates the two groups. We test the accuracy of our approach on a data set consisting of Magnetic Resonance scans from 16 first episode schizophrenics and 17 age-matched healthy control subjects. The data set also contains manual labels for four regions of interest in both hemispheres: superior temporal gyrus, amygdala, hippocampus, and para-hippocampal gyrus. On this small size data set, our approach, which performs classification based on the MR images directly, yields a leave-one-out cross-validation accuracy of up to 90%. This compares favorably with the accuracy achieved by state-of-the-art techniques in schizophrenia MRI research.
This paper presents a method for correcting the geometric and greyscale distortions in diffusion-weighted MRI that result from inhomogeneities in the static magnetic field. These inhomogeneities may due to imperfections in the magnet or to spatial variations in the magnetic susceptibility of the object being imaged—so called susceptibility artifacts. Echo-planar imaging (EPI), used in virtually all diffusion weighted acquisition protocols, assumes a homogeneous static field, which generally does not hold for head MRI. The resulting distortions are significant, sometimes more than ten millimeters. These artifacts impede accurate alignment of diffusion images with structural MRI, and are generally considered an obstacle to the joint analysis of connectivity and structure in head MRI. In principle, susceptibility artifacts can be corrected by acquiring (and applying) a field map. However, as shown in the literature and demonstrated in this paper, field map corrections of susceptibility artifacts are not entirely accurate and reliable, and thus field maps do not produce reliable alignment of EPIs with corresponding structural images. This paper presents a new, image-based method for correcting susceptibility artifacts. The method relies on a variational formulation of the match between an EPI baseline image and a corresponding T2-weighted structural image but also specifically accounts for the physics of susceptibility artifacts. We derive a set of partial differential equations associated with the optimization, describe the numerical methods for solving these equations, and present results that demonstrate the effectiveness of the proposed method compared with field-map correction.
Risholm P, Samsett E, Talos IF, Wells WM III. A Non-rigid Registration Framework that Accommodates Resection and Retraction. Inf Process Med Imaging. 2009;21:447–58.
Traditional non-rigid registration algorithms are incapable of accurately registering intra-operative with pre-operative images whenever tissue has been resected or retracted. In this work we present methods for detecting and handling retraction and resection. The registration framework is based on the bijective Demons algorithm using an anisotropic diffusion smoother. Retraction is detected at areas of the deformation field with high internal strain and the estimated retraction boundary is integrated as a diffusion boundary in the smoother to allow discontinuities to develop across the resection boundary. Resection is detected by a level set method evolving in the space where image intensities disagree. The estimated resection is integrated into the smoother as a diffusion sink to restrict image forces originating inside the resection from being diffused to surrounding areas. In addition, the deformation field is continuous across the diffusion sink boundary which allow us to move the boundary of the diffusion sink without changing values in the deformation field (no interpolation or extrapolation is needed). We present preliminary results on both synthetic and clinical data which clearly shows the added value of explicitly modeling these processes in a registration framework.
Sampat MP, Berger AM, Healy BC, Hildenbrand P, Vass J, Meier DS, Chitnis T, Weiner HL, Bakshi R, Guttmann CRG. Regional white matter atrophy—based classification of multiple sclerosis in cross-sectional and longitudinal data. AJNR Am J Neuroradiol. 2009;30(9):1731–9.
BACKGROUND AND PURPOSE: The different clinical subtypes of multiple sclerosis (MS) may reflect underlying differences in affected neuroanatomic regions. Our aim was to analyze the effectiveness of jointly using the inferior subolivary medulla oblongata volume (MOV) and the cross-sectional area of the corpus callosum in distinguishing patients with relapsing-remitting multiple sclerosis (RRMS), secondary-progressive multiple sclerosis (SPMS), and primary-progressive multiple sclerosis (PPMS). MATERIALS AND METHODS: We analyzed a cross-sectional dataset of 64 patients (30 RRMS, 14 SPMS, 20 PPMS) and a separate longitudinal dataset of 25 patients (114 MR imaging examinations). Twelve patients in the longitudinal dataset had converted from RRMS to SPMS. For all images, the MOV and corpus callosum were delineated manually and the corpus callosum was parcellated into 5 segments. Patients from the cross-sectional dataset were classified as RRMS, SPMS, or PPMS by using a decision tree algorithm with the following input features: brain parenchymal fraction, age, disease duration, MOV, total corpus callosum area and areas of 5 segments of the corpus callosum. To test the robustness of the classification technique, we applied the results derived from the cross-sectional analysis to the longitudinal dataset. RESULTS: MOV and central corpus callosum segment area were the 2 features retained by the decision tree. Patients with MOV >0.94 cm(3) were classified as having RRMS. Patients with progressive MS were further subclassified as having SPMS if the central corpus callosum segment area was
Toews M, Wells WM III. Bayesian Registration via Local Image Regions: Information, Selection and Marginalization. Inf Process Med Imaging. 2009;21:435–46.
We propose a novel Bayesian registration formulation in which image location is represented as a latent random variable. Location is marginalized to determine the maximum a priori (MAP) transform between images, which results in registration that is more robust than the alternatives of omitting locality (i.e. global registration) or jointly maximizing locality and transform (i.e. iconic registration). A mathematical link is established between the Bayesian registration formulation and the mutual information (MI) similarity measure. This leads to a novel technique for selecting informative image regions for registration, based on the MI of image intensity and spatial location. Experimental results demonstrate the effectiveness of the marginalization formulation and the MI-based region selection technique for ultrasound (US) to magnetic resonance (MR) registration in an image-guided neurosurgical application.
Lashkari D, Golland P. Exploratory fMRI analysis without spatial normalization. Inf Process Med Imaging. 2009;21:398–410.
We present an exploratory method for simultaneous parcellation of multisubject fMRI data into functionally coherent areas. The method is based on a solely functional representation of the fMRI data and a hierarchical probabilistic model that accounts for both intersubject and intra-subject forms of variability in fMRI response. We employ a Variational Bayes approximation to fit the model to the data. The resulting algorithm finds a functional parcellation of the individual brains along with a set of population-level clusters, establishing correspondence between these two levels. The model eliminates the need for spatial normalization while still enabling us to fuse data from several subjects. We demonstrate the application of our method on a visual fMRI study.
BACKGROUND: Various osteotomy techniques have been developed to correct the deformity caused by slipped capital femoral epiphysis (SCFE) and compared by their clinical outcomes. The aim of the presented study was to compare an intertrochanteric uniplanar flexion osteotomy with a multiplanar osteotomy by their ability to improve postoperative range of motion as measured by simulation of computed tomographic data in patients with SCFE. METHODS: We examined 19 patients with moderate or severe SCFE as classified based on slippage angle. A computer program for the simulation of movement and osteotomy developed in our laboratory was used for study execution. According to a 3-dimensional reconstruction of the computed tomographic data, the physiological range was determined by flexion, abduction, and internal rotation. The multiplanar osteotomy was compared with the uniplanar flexion osteotomy. Both intertrochanteric osteotomy techniques were simulated, and the improvements of the movement range were assessed and compared. RESULTS: The mean slipping and thus correction angles measured were 25 degrees (range, 8-46 degrees) inferior and 54 degrees (range, 32-78 degrees) posterior. After the simulation of multiplanar osteotomy, the virtually measured ranges of motion as determined by bone-to-bone contact were 61 degrees for flexion, 57 degrees for abduction, and 66 degrees for internal rotation. The simulation of the uniplanar flexion osteotomy achieved a flexion of 63 degrees, an abduction of 36 degrees, and an internal rotation of 54 degrees. CONCLUSIONS: Apart from abduction, the improvement in the range of motion by a uniplanar flexion osteotomy is comparable with that of the multiplanar osteotomy. However, the improvement in flexion for the simulation of both techniques is not satisfactory with regard to the requirements of normal everyday life, in contrast to abduction and internal rotation. LEVEL OF EVIDENCE: Level III, Retrospective comparative study.
Wang X, Grimson EL, Westin CF. Tractography segmentation using a hierarchical Dirichlet processes mixture model. Inf Process Med Imaging. 2009;21:101–13.
In this paper, we propose a new nonparametric Bayesian framework to cluster white matter fiber tracts into bundles using a hierarchical Dirichlet processes mixture (HDPM) model. The number of clusters is automatically learnt from data with a Dirichlet process (DP) prior instead of being manually specified. After the models of bundles have been learnt from training data without supervision, they can be used as priors to cluster/classify fibers of new subjects. When clustering fibers of new subjects, new clusters can be created for structures not observed in the training data. Our approach does not require computing pairwise distances between fibers and can cluster a huge set of fibers across multiple subjects without subsampling. We present results on multiple data sets, the largest of which has more than 120, 000 fibers.