This paper presents a patch based method to normalize temporal intensities from longitudinal brain magnetic resonance (MR) images. Longitudinal intensity normalization is relevant for subsequent processing, such as segmentation, so that rates of change of tissue volumes, cortical thickness, or shapes of brain structures becomes stable and smooth over time. Instead of using intensities at each voxel, we use patches as image features as a patch encodes neighborhood information of the center voxel. Once all the time-points of a longitudinal dataset are registered, the longitudinal intensity change at each patch is assumed to follow an auto-regressive (AR(1)) process. An estimate of the normalized intensities of a patch at every time-point are generated from a hidden Markov model, where the hidden states are the unobserved normalized patches and the outputs are the observed patches. A validation study on a phantom dataset shows good segmentation overlap with the truth, and an experiment with real data shows more stable rates of change for tissue volumes with the temporal normalization than without.
Intensity normalization; intensity standardization; MRI; patch; brain
Deformable registration techniques play vital roles in a variety of medical imaging tasks such as image fusion, segmentation, and post-operative surgery assessment. In recent years, mutual information has become one of the most widely used similarity metrics for medical image registration algorithms. Unfortunately, as a matching criteria, mutual information loses much of its effectiveness when there is poor statistical consistency and a lack of structure. This is especially true in areas of images where the intensity is homogeneous and information is sparse. Here we present a method designed to address this problem by integrating distance transforms of anatomical segmentations as part of a multi-channel mutual information framework within the registration algorithm. Our method was tested by registering real MR brain data and comparing the segmentation of the results against that of the target. Our analysis showed that by integrating distance transforms of the the white matter segmentation into the registration, the overall segmentation of the registration result was closer to the target than when the distance transform was not used.
Image registration; Magnetic resonance imaging; Multidimensional signal processing; Spatial normalization; Distance Transform
Magnetic resonance (MR) images of the tongue have been used in both clinical studies and scientific research to reveal tongue structure. In order to extract different features of the tongue and its relation to the vocal tract, it is beneficial to acquire three orthogonal image volumes—e.g., axial, sagittal, and coronal volumes. In order to maintain both low noise and high visual detail and minimize the blurred effect due to involuntary motion artifacts, each set of images is acquired with an in-plane resolution that is much better than the through-plane resolution. As a result, any one data set, by itself, is not ideal for automatic volumetric analyses such as segmentation, registration, and atlas building or even for visualization when oblique slices are required. This paper presents a method of super-resolution volume reconstruction of the tongue that generates an isotropic image volume using the three orthogonal image volumes. The method uses preprocessing steps that include registration and intensity matching and a data combination approach with the edge-preserving property carried out by Markov random field optimization. The performance of the proposed method was demonstrated on fifteen clinical datasets, preserving anatomical details and yielding superior results when compared with different reconstruction methods as visually and quantitatively assessed.
Super-resolution volume reconstruction; human tongue; magnetic resonance imaging (MRI)
Deformable models are widely used for image segmentation, most commonly to find single objects within an image. Although several methods have been proposed to segment multiple objects using deformable models, substantial limitations in their utility remain. This paper presents a multiple object segmentation method using a novel and efficient object representation for both two and three dimensions. The new framework guarantees object relationships and topology, prevents overlaps and gaps, enables boundary-specific speeds, and has a computationally efficient evolution scheme that is largely independent of the number of objects. Maintaining object relationships and straightforward use of object-specific and boundary-specific smoothing and advection forces enables the segmentation of objects with multiple compartments, a critical capability in the parcellation of organs in medical imaging. Comparing the new framework with previous approaches shows its superior performance and scalability.
Multiple object segmentation; geometric deformable model; level sets; topology preservation
Measuring the internal muscular motion and deformation of the tongue during natural human speech is of high interest to head and neck surgeons and speech language pathologists. A pipeline for calculating 3D tongue motion from dynamic cine and tagged Magnetic Resonance (MR) images during speech has been developed. This paper presents the result of a complete analysis of eleven subjects’ (seven normal controls and four glossectomy patients) global tongue motion during speech obtained through MR imaging and processed through the tongue motion analysis pipeline. The data is regularized into the same framework for comparison. A generalized two-step principal component analysis is used to show the major difference between patients’ and controls’ tongue motions. A test is performed to demonstrate the ability of this process to distinguish patient data from control data and to show the potential power of quantitative analysis that the tongue motion pipeline can achieve.
Tongue; motion; glossectomy; MRI; tagged; HARP; IDEA algorithm; PCA
Accurate segmentation is an important preprocessing step for measuring the internal deformation of the tongue during speech and swallowing using 3D dynamic MRI. In an MRI stack, manual segmentation of every 2D slice and time frame is time-consuming due to the large number of volumes captured over the entire task cycle. In this paper, we propose a semi-automatic segmentation workflow for processing 3D dynamic MRI of the tongue. The steps comprise seeding a few slices, seed propagation by deformable registration, random walker segmentation of the temporal stack of images and 3D super-resolution volumes. This method was validated on the tongue of two subjects carrying out the same speech task with multi-slice 2D dynamic cine-MR images obtained at three orthogonal orientations and 26 time frames. The resulting semi-automatic segmentations of 52 volumes showed an average dice similarity coefficient (DSC) score of 0.9 with reduced segmented volume variability compared to manual segmentations.
Tongue; segmentation; random walker; deformable registration; super-resolution reconstruction
Magnetic resonance imaging (MRI) is widely used for analyzing human brain structure and function. MRI is extremely versatile and can produce different tissue contrasts as required by the study design. For reasons such as patient comfort, cost, and improving technology, certain tissue contrasts for a cohort analysis may not have been acquired during the imaging session. This missing pulse sequence hampers consistent neuroanatomy research. One possible solution is to synthesize the missing sequence. This paper proposes a data-driven approach to image synthesis, which provides equal, if not superior synthesis compared to the state-of-the-art, in addition to being an order of magnitude faster. The synthesis transformation is done on image patches by a trained bagged ensemble of regression trees. Validation was done by synthesizing T2-weighted contrasts from T1-weighted scans, for phantoms and real data. We also synthesized 3 Tesla T1-weighted magnetization prepared rapid gradient echo (MPRAGE) images from 1.5 Tesla MPRAGEs to demonstrate the generality of this approach.
Image synthesis; regression; brain
The superior cerebellar peduncles (SCPs) are white matter tracts that serve as the major efferent pathways from the cerebellum to the thalamus. With diffusion tensor images (DTI), tractography algorithms or volumetric segmentation methods have been able to reconstruct part of the SCPs. However, when the fibers cross, the primary eigenvector (PEV) no longer represents the primary diffusion direction. Therefore, at the crossing of the left and right SCP, known as the decussation of the SCPs (dSCP), fiber tracts propagate incorrectly. To our knowledge, previous methods have not been able to segment the SCPs correctly. In this work, we explore the diffusion properties and seek to volumetrically segment the complete SCPs. The non-crossing SCPs and dSCP are modeled as different objects. A multi-object geometric deformable model is employed to define the boundaries of each piece of the SCPs, with the forces derived from diffusion properties as well as the PEV. We tested our method on a software phantom and real subjects. Results indicate that our method is able to the resolve the crossing and segment the complete SCPs with repeatability.
SCP; MGDM; GGVF; Westin index; Fiber crossing
Magnetic resonance (MR) imaging (MRI) is widely used to study the structure of human brains. Unlike computed tomography (CT), MR image intensities do not have a tissue specific interpretation. Thus images of the same subject obtained with either the same imaging sequence on different scanners or with differing parameters have widely varying intensity scales. This inconsistency introduces errors in segmentation, and other image processing tasks, thus necessitating image intensity standardization. Compared to previous intensity normalization methods using histogram transformations–which try to find a global one-to-one intensity mapping based on histograms–we propose a patch based generative model for intensity normalization between images acquired under different scanners or different pulse sequence parameters. Our method outperforms histogram based methods when normalizing phantoms simulated with various parameters. Additionally, experiments on real data, acquired under a variety of scanners and acquisition parameters, have more consistent segmentations after our normalization.
MRI; intensity normalization; intensity standardization; brain; segmentation
Optical coherence tomography (OCT) of the macular cube has become an increasingly important tool for investigating and managing retinal pathology. One important new area of investigation is the analysis of anatomic variably across a population. Such an analysis on the retina requires the construction of a normalized space, which is generally created through deformable registration of each subject into a common template. Unfortunately, state-of-the-art 3D registration tools fail to adequately spatially normalize retinal OCT images. This work proposes a new deformable registration algorithm for OCT images using the similarity between pairs of A-mode scans. First, a retinal OCT specific affine step is presented, which uses automated landmarks to perform global translations and individual rescaling of all the subject’s A-mode scans. Then, a deformable registration using regularized one-dimensional radial basis functions is applied to further align the retinal layers. Results on 15 subjects show the improved accuracy of this approach in comparison to state of the art methods with respect to registration for labeling. Additional results show the ability to generate stereotaxic spaces for retinal OCT.
Optical coherence tomography; registration
Accurate localization of myocardial viability is important in diagnosis of infarction. Regional strain function provides excessive information for clinical decision making but comparison of strain tensor profiles across differing tissue types is usually difficult due to multivariate nature of tensors. It is desirable to describe tensors with simplified scalar indices which are more mathematically and statistically intuitive. In this work, anisotropy of tensors in healthy and experimental infarct regions in a large animal model is assessed and compared to directional components of strain tensors which are currently the most popular indices in active use. Myocardial strain tensors are computed using zHARP, a magnetic resonance (MR) tagging technique that provides quantification of cardiac function with direct computation of three-dimensional tensors from two-dimensional short axis MR images. Fractional anisotropy of strain tensors shows high correlation with late gadolinium enhanced images and is capable of discrimination between healthy and infarcted regions.
Strain tensor; HARP; zHARP; Fractional Anisotropy; Infarction
We present a compressed sensing based approach to remove gain field from magnetic resonance (MR) images of the human brain. During image acquisition, the inhomogeneity present in the radio-frequency (RF) coil appears as shading artifact in the intensity image. The inhomogeneity poses problem in any automatic algorithm that uses intensity as a feature. It has been shown that at low field strength, the shading can be assumed to be a smooth field that is composed of low frequency components. Thus most inhomogeneity correction algorithms assume some kind of explicit smoothness criteria on the field. This sometimes limits the performance of the algorithms if the actual inhomogeneity is not smooth, which is the case at higher field strength. We describe a model-free, non-parametric patch-based approach that uses compressed sensing for the correction. We show that these features enable our algorithm to perform comparably with a current state of the art method N3 on images acquired at low field, while outperforming N3 when the image has non-smooth inhomogeneity, such as 7T images.
MRI; intensity non-uniformity; intensity inhomogeneity; 7T; bias field; bias correction
Volumetric measurements obtained from image parcellation have been instrumental in uncovering structure-function relationships. However, anatomical study of the cerebellum is a challenging task. Because of its complex structure, expert human raters have been necessary for reliable and accurate segmentation and parcellation. Such delineations are time-consuming and prohibitively expensive for large studies. Therefore, we present a three-part cerebellar parcellation system that utilizes multiple inexpert human raters that can efficiently and expediently produce results nearly on par with those of experts. This system includes a hierarchical delineation protocol, a rapid verification and evaluation process, and statistical fusion of the inexpert rater parcellations. The quality of the raters’ and fused parcellations was established by examining their Dice similarity coefficient, region of interest (ROI) volumes, and the intraclass correlation coefficient of region volume. The intra-rater ICC was found to be 0.93 at the finest level of parcellation.
Human Cerebellum; Manual labeling; Delineation; Parcellation; STAPLE; STAPLER; Label fusion
Intensity normalization is an important preprocessing step in magnetic resonance (MR) image analysis. In MR images (MRI), the observed intensities are primarily dependent on (1) intrinsic magnetic resonance properties of the tissues such as proton density (PD), longitudinal and transverse relaxation times (T1 and T2 respectively), and (2) the scanner imaging parameters like echo time (TE), repeat time (TR), and flip angle (α). We propose a method which utilizes three co-registered images with different contrast mechanisms (PD-weighted, T2-weighted and T1-weighted) to first estimate the imaging parameters and then estimate PD, T1, and T2 values. We then normalize the subject intensities to a reference by simply applying the pulse sequence equation of the reference image to the subject tissue parameters. Previous approaches to solve this problem have primarily focused on matching the intensity histograms of the subject image to a reference histogram by different methods. The fundamental drawback of these methods is their failure to respect the underlying imaging physics and tissue biology. Our method is validated on phantoms and we show improvement of normalization on real images of human brains.
intensity normalization/standardization; brain; magnetic resonance imaging; pulse sequence
With the rapid development of microscopy for cell imaging, there is a strong and growing demand for image analysis software to quantitatively study cell morphology. Automatic cell segmentation is an important step in image analysis. Despite substantial progress, there is still a need to improve the accuracy, efficiency, and adaptability to different cell morphologies. In this paper, we propose a fully automatic method for segmenting cells in fluorescence images of confluent cell monolayers. This method addresses several challenges through a combination of ideas. 1) It realizes a fully automatic segmentation process by first detecting the cell nuclei as initial seeds and then using a multi-object geometric deformable model (MGDM) for final segmentation. 2) To deal with different defects in the fluorescence images, the cell junctions are enhanced by applying an order-statistic filter and principal curvature based image operator. 3) The final segmentation using MGDM promotes robust and accurate segmentation results, and guarantees no overlaps and gaps between neighboring cells. The automatic segmentation results are compared with manually delineated cells, and the average Dice coefficient over all distinguishable cells is 0.88.
Cell segmentation; immunofluorescence microscopy; cell nuclei; cell junction network; multi-object geometric deformable model (MGDM)
The lack of dynamic dosimetry tools for permanent prostate brachytherapy causes otherwise avoidable problems in prostate cancer patient care. The goal of this work is to satisfy this need in a readily adoptable manner. Using the ubiquitous ultrasound scanner and mobile non-isocentric C-arm, we show that dynamic dosimetry is now possible with only the addition of an arbitrarily configured marker-based fiducial. Not only is the system easily configured from accessible hardware, but it is also simple and convenient, requiring little training from technicians. Furthermore, the proposed system is built upon robust algorithms of seed segmentation, fiducial detection, seed reconstruction, and image registration. All individual steps of the pipeline have been thoroughly tested, and the system as a whole has been validated on a study of 25 patients. The system has shown excellent results of accurately computing dose, and does so with minimal manual intervention, therefore showing promise for widespread adoption of dynamic dosimetry.
prostate brachytherapy; dynamic dosimetry; mobile non-isocentric C-arm; transrectal ultrasound
The thalamus is a sub-cortical gray matter structure that relays signals between the cerebral cortex and midbrain. It can be parcellated into the thalamic nuclei which project to different cortical regions. The ability to automatically parcellate the thalamic nuclei could lead to enhanced diagnosis or prognosis in patients with some brain disease. Previous works have used diffusion tensor images (DTI) to parcellate the thalamus, using either tensor similarity or cortical connectivity as information driving the parcellation. In this paper, we propose a method that uses the diffusion tensors in a different way than previous works to guide a multiple object geometric deformable model (MGDM) for parcellation. The primary eigenvector (PEV) is used to indicate the homogeneity of fiber orientations. To remove the ambiguity due to the fact that the PEV is an orientation, we map the PEV into a 5D space known as the Knutsson space. An edge map is then generated from the 5D vector to show divisions between regions of aligned PEV’s. The generalized gradient vector flow (GGVF) calculated from the edge map drives the evolution of the boundary of each nucleus. Region based force, balloon force, and curvature force are also employed to refine the boundaries. Experiments have been carried out on five real subjects. Quantitative measures show that the automated parcellation agrees with the manual delineation of an expert under a published protocol.
thalamic parcellation; DTI; 5D Knutsson space; multiple object geometric deformable model
Active shape models (ASMs) have been widely used in segmentation tasks in medical image analysis. Complex structures and a limited number of training samples can, however, result in the failure to capture the complete range of shape variations. Various modifications to the point distribution model (PDM) have been proposed to increase the flexibility of the model. Still model parameters are often determined empirically without respect to the underlying data structure. We explore shrinkage covariance estimation in building a PDM by combining the sample covariance matrix with a target covariance matrix estimated from a low-dimensional constrained model. Instead of using a global shrinkage intensity, we apply a spatially varying shrinkage intensity field to better adapt to the spatially varying characteristic of a complex shape. The parameters of the constrained model and the amount of shrinkage are determined in a data-driven fashion, so that the resulting distribution is optimized in representing the underlying data. The PDM, which we call SC-PDM, shows an increased flexibility in fitting new shapes and at the same time, is robust to noise. We demonstrate the effectiveness of using SC-PDM to label gyral regions on the human cerebral cortex.
Active shape model; point distribution model; covariance shrinking; cerebral cortex; gyral labeling
While neurodegenerative diseases are characterized by steady degeneration over relatively long timelines, it is widely believed that the early stages are the most promising for therapeutic intervention, before irreversible neuronal loss occurs. Developing a therapeutic response requires a precise measure of disease progression. However, since the early stages are for the most part asymptomatic, obtaining accurate measures of disease progression is difficult. Longitudinal databases of hundreds of subjects observed during several years with tens of validated biomarkers are becoming available, allowing the use of computational methods. We propose a widely applicable statistical methodology for creating a disease progression score (DPS), using multiple biomarkers, for subjects with a neurodegenerative disease. The proposed methodology was evaluated for Alzheimer’s disease (AD) using the publicly available AD Neuroimaging Initiative (ADNI) database, yielding an Alzheimer’s DPS or ADPS score for each subject and each time-point in the database. In addition, a common description of biomarker changes was produced allowing for an ordering of the biomarkers. The Rey Auditory Verbal Learning Test delayed recall was found to be the earliest biomarker to become abnormal. The group of biomarkers comprising the volume of the hippocampus and the protein concentration amyloid beta and Tau were next in the timeline, and these were followed by three cognitive biomarkers. The proposed methodology thus has potential to stage individuals according to their state of disease progression relative to a population and to deduce common behaviors of biomarkers in the disease itself.
Neurodegenerative diseases; Alzheimer’s disease; biomarkers; disease progression score
We present a method that utilizes registration displacement fields to perform accurate classification of magnetic resonance images (MRI) of the brain acquired from healthy individuals and patients diagnosed with multiple sclerosis (MS). Contrary to standard approaches, each voxel in the displacement field is treated as an independent feature that is classified individually. Results show that when used with a simple linear discriminant and majority voting, the approach is superior to using the displacement field with a single classifier, even when compared against more sophisticated classification methods such as adaptive boosting, random forests, and support vector machines. Leave-one-out cross-validation was used to evaluate this method for classifying images by disease, MS subtype (Acc: 77%–88%), and age (Acc: 96%–100%).
Image registration; Magnetic resonance imaging; Classification; Multiple Sclerosis
Labeling of cerebral vasculature is important for characterization of anatomical variation, quantification of brain morphology with respect to specific vessels, and inter-subject comparisons of vessel properties and abnormalities. We propose an automated method to label the anterior portion of cerebral arteries using a statistical inference method on the Bayesian network representation of the vessel tree. Our approach combines the likelihoods obtained from a random forest classifier trained using vessel centerline features with a belief propagation method integrating the connection probabilities of the cerebral artery network. We evaluate our method on 30 subjects using a leave-one-out validation, and show that it achieves an average correct vessel labeling rate of over 92%.
Automated labeling of vessels; cerebral arteries; random forest; belief propagation; statistical inference on Bayesian networks
To understand the role of the tongue in speech production, it is desirable to directly image the motion and strain of the muscles within the tongue. Magnetic resonance tagging—which was originally developed for cardiac imaging—has previously been applied to image both two-dimensional and three-dimensional tongue motion during speech. However, to quantify three-dimensional motion and strain, multiple images yielding two-dimensional motion must be acquired at different orientations and then interpolated—a time-consuming task both in image acquisition and processing. Recently, a new MR imaging and image processing method called zHARP was developed to encode and track 3D motion from a single slice without increasing acquisition time. zHARP was originally developed and applied to cardiac imaging. The application of zHARP to the tongue is not straightforward because the tongue in repetitive speech does not move as consistently as the heart in its beating cycle. Therefore tongue images are more susceptible to motion artifacts. Moreover, these artifacts are greatly exaggerated as compared to conventional tagging because of the nature of zHARP acquisition. In this work, we re-implemented the zHARP imaging sequence and optimized it for the tongue motion analysis. We also optimized image acquisition by designing and developing a specialized MRI scanner triggering method and vocal repetition to better synchronize speech repetitions. Our method was validated using a moving phantom. Results of 3D motion tracking and strain analysis on the tongue experiments demonstrate the effectiveness of this method.
Motion quantification; tongue; zHARP
The thalamus sub-cortical gray matter structure consists of contiguous nuclei, each individually responsible for communication between various cerebral cortex and midbrain regions. These nuclei are differentially affected in neurodegenerative diseases such as multiple sclerosis and Alzheimer’s. However thalamic parcellation of the nuclei, manual or automatic, is difficult given the limited contrast in any particular magnetic resonance (MR) modality. Several groups have had qualitative success differentiating nuclei based on spatial location and fiber orientation information in diffusion tensor imaging (DTI). In this paper, we extend these principles by combining these discriminating dimensions with structural MR and derived information, and by building random forest learners on the resultant multi-modal features. In training, we form a multi-dimensional feature per voxel, which we associate with a nucleus classification from a manual rater. Learners are trained to differentiate thalamus from background and thalamic nuclei from other nuclei. These learners inform the external forces of a multiple object level set model. Our cross-validated quantitative results on a set of twenty subjects show the efficacy and reproducibility of our results.
Diffusion tensor imaging; machine learning; deformable models; object segmentation; random forests
In this study, we used manual delineation of high-resolution magnetic resonance imaging (MRI) to determine the spatial and temporal characteristics of the cerebellar atrophy in spinocerebellar ataxia type 2 (SCA2). Ten subjects with SCA2 were compared to ten controls. The volume of the pons, the total cerebellum, and the individual cerebellar lobules were calculated via manual delineation of structural MRI. SCA2 showed substantial global atrophy of the cerebellum. Furthermore, the degeneration was lobule-specific, selectively affecting the anterior lobe, VI, Crus I, Crus II, VIII, uvula, corpus medullare, and pons, while sparing VIIB, tonsil/paraflocculus, flocculus, declive, tuber/folium, pyramis, and nodulus. The temporal characteristics differed in each cerebellar subregion: 1) Duration of disease: Crus I, VIIB, VIII, uvula, corpus medullare, pons, and the total cerebellar volume correlated with the duration of disease; 2) Age: VI, Crus II, and flocculus correlated with age in control subjects; 3) Clinical scores: VI, Crus I, VIIB, VIII, corpus medullare, pons, and the total cerebellar volume correlated with clinical scores in SCA2. No correlations were found with the age of onset. Our extrapolated volumes at the onset of symptoms suggest that neurodegeneration may be present even during the presymptomatic stages of disease. The spatial and temporal characteristics of the cerebellar degeneration in SCA2 are region-specific. Furthermore, our findings suggest the presence of presymptomatic atrophy and a possible developmental component to the mechanisms of pathogenesis underlying SCA2. Our findings further suggest that volumetric analysis may aid in the development of a non-invasive, quantitative biomarker.
ataxia; spinocerebellar ataxia type 2 (SCA2); magnetic resonance imaging (MRI); biomarker
Tissue contrast and resolution of magnetic resonance neuroimaging data have strong impacts on the utility of the data in clinical and neuroscience tasks such as registration and segmentation. Lengthy acquisition times typically prevent routine acquisition of multiple MR tissue contrast images at high resolution, and the opportunity for detailed analysis using these data would seem to be irrevocably lost. This paper describes an example based approach using patch matching from a multiple resolution multiple contrast atlas in order to change an image's resolution as well as its MR tissue contrast from one pulse-sequence to that of another. The use of this approach to generate different tissue contrasts (T2/PD/FLAIR) from a single T1-weighted image is demonstrated on both phantom and real images.
Image classification; resolution; segmentation; MR tissue contrast; contrast synthesis; image hallucination; atlas