We present a novel multiresolution analysis based stereo matching method using curvelets and modified adaptive support weight. Multiresolution analysis has long been applied to stereo correspondence. However, previous methods suffer from false matches arising from textureless region or repetitive textures and fattening effect due to area based matching. In the proposed approach, we have reduced false matches by using curvelet coefficients in different scales and orientations. Curvelet coefficients can uniquely represent different image points and increase matching accuracy. The fattening effect is reduced using support weights modified for curvelets. The proposed method is verified and compared with state-of-the art methods by extensive tests, and good results are obtained.
This paper introduces a bilateral filtering based mixture model for image segmentation. The mixture model uses Markov Random Field (MRF) to incorporate spatial relationship among neighboring pixels into the Gaussian Mixture Model (GMM) in order to perform a segmentation that is robust against noise and other environmental factors. The bilateral filtering is used to smooth the posterior probability map as part of the MRF used. The advantage of the proposed model is its simplified structure so that the Expectation Maximization algorithm can be directly applied to the log-likelihood function to compute the optimum parameters of the mixture model. The method has been extensively tested on synthetic and natural images and compared with some of the state-of-the-arts algorithms currently available. The experimental results show that the proposed method is comparable to the other methods in terms of accuracy and quality and simpler in terms of implementation.
Abstract Müller glia, the most abundant glia of vertebrate retina, have an elaborate morphology characterized by a vertical stalk that spans the retina and branches in each retinal layer. Müller glia play diverse, critical roles in retinal homeostasis, which are presumably enabled by their complex anatomy. However, much remains unknown, particularly in mouse, about the anatomical arrangement of Müller cells and their arbors, and how these features arise in development. Here we use membrane‐targeted fluorescent proteins to reveal the fine structure of mouse Müller arbors. We find sublayer‐specific arbor specializations within the inner plexiform layer (IPL) that occur consistently at defined laminar locations. We then characterize Müller glia spatial patterning, revealing how individual cells collaborate to form a pan‐retinal network. Müller cells, unlike neurons, are spread across the retina with homogenous density, and their arbor sizes change little with eccentricity. Using Brainbow methods to label neighboring cells in different colors, we find that Müller glia tile retinal space with minimal overlap. The shape of their arbors is irregular but nonrandom, suggesting that local interactions between neighboring cells determine their territories. Finally, we identify a developmental window at postnatal Days 6 to 9 when Müller arbors first colonize the synaptic layers beginning in stereotyped inner plexiform layer sublaminae. Together, our study defines the anatomical arrangement of mouse Müller glia and their network in the radial and tangential planes of the retina, in development and adulthood. The local precision of Müller glia organization suggests that their morphology is sculpted by specific cell to cell interactions with neurons and each other.
Inner retina in Alzheimer's Disease (AD) may experience neuroinflammation resulting in atrophy. The objective of our study was to determine whether retinal GCIPL (ganglion cell-inner plexiform layer) or nerve fiber layer (NFL) thickness may serve as noninvasive biomarkers to diagnose AD. This cross-sectional case-control study enrolled 15 mild cognitive impairment (MCI) patients, 15 mild-moderate AD patients, and 18 cognitively normal adults. NFL and GCIPL thicknesses on optical coherence tomography (OCT) were measured using Duke Optical Coherence Tomography Retinal Analysis Program (DOCTRAP) and Spectralis software. We demonstrated that regional thicknesses of NFL or GCIPL on macular or nerve OCTs did not differ between groups. However, a multi-variate regression analysis identified macular areas with a significant thickening or thinning in NFL and GCIPL in MCI and AD patients. Our primary findings controvert previous reports of thinner NFL in moderate-to-severe AD. The areas of thickening of GCIPL and NFL in the macula adjacent to areas of thinning, as revealed by a more complex statistical model, suggest that NFL and GCIPL may undergo dynamic changes during AD progression.
Segmentation of a medical image based on the modeling and estimation of the tissue intensity probability density functions via a Gaussian mixture model has recently received great attention. However, the Gaussian distribution is unbounded and symmetrical around its mean. This study presents a new bounded asymmetric mixture model for analyzing both univariate and multivariate data. The advantage of the proposed model is that it has the flexibility to fit different shapes of observed data such as non-Gaussian, nonsymmetric, and bounded support data. Another advantage is that each component of the proposed model has the ability to model the observed data with different bounded support regions, which is suitable for application on image segmentation. Our method is intuitively appealing, simple, and easy to implement. We also propose a new method to estimate the model parameters in order to minimize the higher bound on the data negative log-likelihood function. Numerical experiments are presented where the proposed model is tested in various images from simulated to real 3- D medical ones.
Motion segmentation is an important research field as it forms the stepping stone for traffic monitoring, video surveillance, activity analysis, gait recognition and many other automatic imaging applications. In this work, a novel generic multiresolution (MR) based framework has been proposed in conjunction with Sigma-delta based motion segmentation algorithm. The framework provides a general platform to use any MR analysis method to 1) incorporate subbands information containing varying features for enhanced motion extraction and 2) combine the information obtained to incrementally form the background using Sigma-delta method and upscale to original frame resolution. The validity of the proposed method is demonstrated using four popular MR analysis methods. Comparison of the proposed framework with sigma-delta and wavelet based change detection reflects several improvements over these methods.
Visual impairment due to glaucoma currently impacts 70 million people worldwide. While disease progression can be slowed or stopped with effective lowering of intraocular pressure, current medical treatments are often inadequate. Fortunately, three new classes of therapeutics that target the diseased conventional outflow tissue responsible for ocular hypertension are in the final stages of human testing. The rho kinase inhibitors have proven particularly efficacious and additive to current therapies. Unfortunately, non-contact technology that monitors the health of outflow tissue and its response to conventional outflow therapy is not available clinically. Using optical coherence tomographic (OCT) imaging and novel segmentation software, we present the first demonstration of drug effects on conventional outflow tissues in living eyes. Topical netarsudil (formerly AR-13324), a rho kinase/ norepinephrine transporter inhibitor, affected both proximal (trabecular meshwork and Schlemm's Canal) and distal portions (intrascleral vessels) of the mouse conventional outflow tract. Hence, increased perfusion of outflow tissues was reliably resolved by OCT as widening of the trabecular meshwork and significant increases in cross-sectional area of Schlemm's canal following netarsudil treatment. These changes occurred in conjunction with increased outflow facility, increased speckle variance intensity of outflow vessels, increased tracer deposition in conventional outflow tissues and decreased intraocular pressure. This is the first report using live imaging to show real-time drug effects on conventional outflow tissues and specifically the mechanism of action of netarsudil in mouse eyes. Advancements here pave the way for development of a clinic-friendly OCT platform for monitoring glaucoma therapy.
Patient motion artifacts are an important source of data irregularities in OCT imaging. With longer duration OCT scans – as is needed for large wide field of view scans or increased scan density – motion artifacts become increasingly problematic. Strategies to mitigate these motion artifacts are then necessary to ensure OCT data integrity. A popular strategy for reducing motion artifacts in OCT images is to capture two orthogonally oriented volumetric scans containing uncorrelated motion and subsequently reconstructing a motion-free volume by combining information from both datasets. While many different variations of this registration approach have been proposed, even the most recent methods might not be suitable for wide FOV OCT scans which can be lacking in features away from the optic nerve head or arcades. To address this problem, we propose a two-stage motion correction algorithm for wide FOV OCT volumes. In the first step, X and Y axes motion is corrected by registering OCT summed voxel projections (SVPs). To achieve this, we introduce a method based on a custom variation of the dense optical flow technique which is aware of the motion free orientation of the scan. Secondly, a depth (Z axis) correction approach based on the segmentation of the retinal layer boundaries in each B-scan using graph-theory and dynamic programming is applied. This motion correction method was applied to wide field retinal OCT volumes (approximately 80° FOV) of 3 subjects with substantial reduction in motion artifacts.
'%3e%3cpath fill='%23012169' d='M0 0v30h60V0z'/%3e%3cpath stroke='%23fff' stroke-width='6' d='m0 0 60 30m0-30L0 30'/%3e%3cpath stroke='%23C8102E' stroke-width='4' d='m0 0 60 30m0-30L0 30' clip-path='url(%23b)'/%3e%3cpath stroke='%23fff' stroke-width='10' d='M30 0v30M0 15h60'/%3e%3cpath stroke='%23C8102E' stroke-width='6' d='M30 0v30M0 15h60'/%3e%3c/g%3e%3c/svg%3e)
