Mirza Faisal Beg

Measuring and mapping cardiac fiber and laminar architecture using diffusion tensor MR imaging

Abstract: The ventricular myocardium is known to exhibit a complex spatial organization, with fiber orientation varying as a function of transmural location. It is now well established that diffusion tensor magnetic resonance imaging (DTMRI) may be used to measure this fiber orientation at high spatial resolution. Cardiac fibers are also known to be organized in sheets with surface orientation varying throughout the ventricles. This article reviews results on use of DTMRI for measuring ventricular fiber orientation, as well as presents new results providing strong evidence that the tertiary eigenvector of the diffusion tensor is aligned locally with the cardiac sheet surface normal. Considered together, these data indicate that DTMRI may be used to reconstruct both ventricular fiber and sheet organization. This article also presents the large deformation diffeomorphic metric mapping (LDDMM) algorithm and shows that this algorithm may be used to bring ensembles of imaged and reconstructed hearts into correspondence (e.g., registration) so that variability of ventricular geometry, fiber, and sheet orientation may be quantified. Ventricular geometry and fiber structure is known to be remodeled in a range of disease processes; however, descriptions of this remodeling have remained subjective and qualitative. We anticipate that use of DTMRI for reconstruction of ventricular anatomy coupled with application of the LDDMM method for image volume registration will enable the detection and quantification of changes in cardiac anatomy that are characteristic of specific disease processes in the heart. Finally, we show that epicardial electrical mapping and DTMRI imaging may be performed in the same hearts. The anatomic data may then be used to simulate electrical conduction in a computational model of the very same heart that was mapped electrically. This facilitates direct comparison and testing of model versus experimental results and opens the door to quantitative measurement, modeling, and analysis of the ways in which remodeling of ventricular microanatomy influences electrical conduction in the heart.

Evidence of Structural Remodeling in the Dyssynchronous Failing Heart

Ventricular remodeling of both geometry and fiber structure is a prominent feature of several cardiac pathologies. Advances in MRI and analytical methods now make it possible to measure changes of cardiac geometry, fiber, and sheet orientation at high spatial resolution. In this report, we use diffusion tensor imaging to measure the geometry, fiber, and sheet architecture of eight normal and five dyssynchronous failing canine hearts, which were explanted and fixed in an unloaded state. We apply novel computational methods to identify statistically significant changes of cardiac anatomic structure in the failing and control heart populations. The results demonstrate significant regional differences in geometric remodeling in the dyssynchronous failing heart versus control. Ventricular chamber dilatation and reduction in wall thickness in septal and some posterior and anterior regions are observed. Primary fiber orientation showed no significant change. However, this result coupled with the local wall thinning in the septum implies an altered transmural fiber gradient. Further, we observe that orientation of laminar sheets become more vertical in the early-activated septum, with no significant change of sheet orientation in the late-activated lateral wall. Measured changes in both fiber gradient and sheet structure will affect both the heterogeneity of passive myocardial properties as well as electrical activation of the ventricles.

FreeSurfer-Initiated Fully-Automated Subcortical Brain Segmentation in MRI Using Large Deformation Diffeomorphic Metric Mapping

Fully-automated brain segmentation methods have not been widely adopted for clinical use because of issues related to reliability, accuracy, and limitations of delineation protocol. By combining the probabilistic-based FreeSurfer (FS) method with the Large Deformation Diffeomorphic Metric Mapping (LDDMM)-based label-propagation method, we are able to increase reliability and accuracy, and allow for flexibility in template choice. Our method uses the automated FreeSurfer subcortical labeling to provide a coarse-to-fine introduction of information in the LDDMM template-based segmentation resulting in a fully-automated subcortical brain segmentation method (FS+LDDMM). One major advantage of the FS+LDDMM-based approach is that the automatically generated segmentations generated are inherently smooth, thus subsequent steps in shape analysis can directly follow without manual post-processing or loss of detail. We have evaluated our new FS+LDDMM method on several databases containing a total of 50 subjects with different pathologies, scan sequences and manual delineation protocols for labeling the basal ganglia, thalamus, and hippocampus. In healthy controls we report Dice overlap measures of 0.81, 0.83, 0.74, 0.86 and 0.75 for the right caudate nucleus, putamen, pallidum, thalamus and hippocampus respectively. We also find statistically significant improvement of accuracy in FS+LDDMM over FreeSurfer for the caudate nucleus and putamen of Huntingtons disease and Tourettes syndrome subjects, and the right hippocampus of Schizophrenia subjects.

Neonatal Pain-Related Stress Predicts Cortical Thickness at Age 7 Years in Children Born Very Preterm

Background Altered brain development is evident in children born very preterm (24–32 weeks gestational age), including reduction in gray and white matter volumes, and thinner cortex, from infancy to adolescence compared to term-born peers. However, many questions remain regarding the etiology. Infants born very preterm are exposed to repeated procedural pain-related stress during a period of very rapid brain development. In this vulnerable population, we have previously found that neonatal pain-related stress is associated with atypical brain development from birth to term-equivalent age. Our present aim was to evaluate whether neonatal pain-related stress (adjusted for clinical confounders of prematurity) is associated with altered cortical thickness in very preterm children at school age. Methods 42 right-handed children born very preterm (24–32 weeks gestational age) followed longitudinally from birth underwent 3-D T1 MRI neuroimaging at mean age 7.9 yrs. Children with severe brain injury and major motor/sensory/cognitive impairment were excluded. Regional cortical thickness was calculated using custom developed software utilizing FreeSurfer segmentation data. The association between neonatal pain-related stress (defined as the number of skin-breaking procedures) accounting for clinical confounders (gestational age, illness severity, infection, mechanical ventilation, surgeries, and morphine exposure), was examined in relation to cortical thickness using constrained principal component analysis followed by generalized linear modeling. Results After correcting for multiple comparisons and adjusting for neonatal clinical factors, greater neonatal pain-related stress was associated with significantly thinner cortex in 21/66 cerebral regions (p-values ranged from 0.00001 to 0.014), predominately in the frontal and parietal lobes. Conclusions In very preterm children without major sensory, motor or cognitive impairments, neonatal pain-related stress appears to be associated with thinner cortex in multiple regions at school age, independent of other neonatal risk factors.

Computational Cardiac Anatomy Using MRI

Ventricular geometry and fiber orientation may undergo global or local remodeling in cardiac disease. However, there are as yet no mathematical and computational methods for quantifying variation of geometry and fiber orientation or the nature of their remodeling in disease. Toward this goal, a landmark and image intensity‐based large deformation diffeomorphic metric mapping (LDDMM) method to transform heart geometry into common coordinates for quantification of shape and form was developed. Two automated landmark placement methods for modeling tissue deformations expected in different cardiac pathologies are presented. The transformations, computed using the combined use of landmarks and image intensities, yields high‐registration accuracy of heart anatomies even in the presence of significant variation of cardiac shape and form. Once heart anatomies have been registered, properties of tissue geometry and cardiac fiber orientation in corresponding regions of different hearts may be quantified. Magn Reson Med 52:1167–1174, 2004. Published 2004 Wiley‐Liss, Inc.

Symmetric Data Attachment Terms for Large Deformation Image Registration

Nonrigid medical image registration between images that are linked by an invertible transformation is an inherently symmetric problem. The transformation that registers the image pair should ideally be the inverse of the transformation that registers the pair with the order of images interchanged. This property is referred to as symmetry in registration or inverse consistent registration. However, in practical estimation, the available registration algorithms have tended to produce inverse inconsistent transformations when the template and target images are interchanged. In this paper, we propose two novel cost functions in the large deformation diffeomorphic framework that are inverse consistent. These cost functions have symmetric data-attachment terms; in the first, the matching error is measured at all points along the flow between template and target, and in the second, matching is enforced only at the midpoint of the flow between the template and target. We have implemented these cost functions and present experimental results to validate their inverse consistent property and registration accuracy.

Rapid volumetric OCT image acquisition using compressive sampling.

Acquiring three dimensional image volumes with techniques such as Optical Coherence Tomography (OCT) relies on reconstructing the tissue layers based on reflection of light from tissue interfaces. One B-mode scan in an image is acquired by scanning and concatenating several A-mode scans, and several contiguous slices are acquired to assemble a full 3D image volume. In this work, we demonstrate how Compressive Sampling (CS) can be used to accurately reconstruct 3D OCT images with minimal quality degradation from a subset of the original image. The full 3D image is reconstructed from sparsely sampled data by exploiting the sparsity of the image in a carefully chosen transform domain. We use several sub-sampling schemes, recover the full 3D image using CS, and show that there is negligible effect on clinically relevant morphometric measurements of the optic nerve head in the recovered image. The potential outcome of this work is a significant reduction in OCT image acquisition time, with possible extensions to speeding up acquisition in other imaging modalities such as ultrasound and MRI.

Automated detection of amnestic mild cognitive impairment in community-dwelling elderly adults: a combined spatial atrophy and white matter alteration approach.

Amnestic mild cognitive impairment (aMCI) is a syndrome widely considered to be prodromal Alzheimers disease. Accurate diagnosis of aMCI would enable earlier treatment, and could thus help minimize the prevalence of Alzheimers disease. The aim of the present study was to evaluate a magnetic resonance imaging-based automated classification schema for identifying aMCI. This was carried out in a sample of community-dwelling adults aged 70-90 years old: 79 with a clinical diagnosis of aMCI and 204 who were cognitively normal. Our schema was novel in using measures of both spatial atrophy, derived from T1-weighted images, and white matter alterations, assessed with diffusion tensor imaging (DTI) tract-based spatial statistics (TBSS). Subcortical volumetric features were extracted using a FreeSurfer-initialized Large Deformation Diffeomorphic Metric Mapping (FS+LDDMM) segmentation approach, and fractional anisotropy (FA) values obtained for white matter regions of interest. Features were ranked by their ability to discriminate between aMCI and normal cognition, and a support vector machine (SVM) selected an optimal feature subset that was used to train SVM classifiers. As evaluated via 10-fold cross-validation, the classification performance characteristics achieved by our schema were: accuracy, 71.09%; sensitivity, 51.96%; specificity, 78.40%; and area under the curve, 0.7003. Additionally, we identified numerous socio-demographic, lifestyle, health and other factors potentially implicated in the misclassification of individuals by our schema and those previously used by others. Given its high level of performance, our classification schema could facilitate the early detection of aMCI in community-dwelling elderly adults.

Real-time high-speed volumetric imaging using compressive sampling optical coherence tomography

Volumetric imaging of the Optic Nerve Head (ONH) morphometry with Optical Coherence Tomography (OCT) requires dense sampling and relatively long acquisition times. Compressive Sampling (CS) is an emerging technique to reduce volume acquisition time with minimal image degradation by sparsely sampling the object and reconstructing the missing data in software. In this report, we demonstrated real-time CS-OCT for volumetric imaging of the ONH using a 1060nm Swept-Source OCT prototype. We also showed that registration and averaging of CS-recovered volumes enhanced visualization of deep structures of the sclera and lamina cribrosa. This work validates CS-OCT as a means for reducing volume acquisition time and for preserving high-resolution in volume-averaged images. Compressive sampling can be integrated into new and existing OCT systems without changes to the optics, requiring only software changes and post-processing of acquired data.

A semi-automated method for identifying and measuring myelinated nerve fibers in scanning electron microscope images

Diagnosing illnesses, developing and comparing treatment methods, and conducting research on the organization of the peripheral nervous system often require the analysis of peripheral nerve images to quantify the number, myelination, and size of axons in a nerve. Current methods that require manually labeling each axon can be extremely time-consuming as a single nerve can contain thousands of axons. To improve efficiency, we developed a computer-assisted axon identification and analysis method that is capable of analyzing and measuring sub-images covering the nerve cross-section, acquired using a scanning electron microscope. This algorithm performs three main procedures - it first uses cross-correlation to combine the acquired sub-images into a large image showing the entire nerve cross-section, then identifies and individually labels axons using a series of image intensity and shape criteria, and finally identifies and labels the myelin sheath of each axon using a region growing algorithm with the geometric centers of axons as seeds. To ensure accurate analysis of the image, we incorporated manual supervision to remove mislabeled axons and add missed axons. The typical user-assisted processing time for a two-megapixel image containing over 2000 axons was less than 1h. This speed was almost eight times faster than the time required to manually process the same image. Our method has proven to be well suited for identifying axons and their characteristics, and represents a significant time savings over traditional manual methods.