A5024783660- Video Surveillance and Tracking Methods
- Advanced Vision and Imaging
- Bayesian Methods and Mixture Models
- Retinal Imaging and Analysis
- Retinal Diseases and Treatments
- Video Analysis and Summarization
- Image Enhancement Techniques
- Advanced Image and Video Retrieval Techniques
- Glaucoma and retinal disorders
- Image Retrieval and Classification Techniques
- Retinal Development and Disorders
- Neuroscience and Neuropharmacology Research
- Fire Detection and Safety Systems
- Retinal and Macular Surgery
- Robotics and Sensor-Based Localization
- Medical Image Segmentation Techniques
- Retinopathy of Prematurity Studies
- Advanced Image Fusion Techniques
- Retinal and Optic Conditions
- Advanced Clustering Algorithms Research
- Human Pose and Action Recognition
- Colorectal Cancer Screening and Detection
- Advanced Image Processing Techniques
- Machine Learning and ELM
- Protein Degradation and Inhibitors
National Institute of Technology Durgapur
2023
Duke University
2015-2022
Epson (United States)
2019
Duke Medical Center
2017-2019
Pratt Institute
2017
University of Windsor
2010-2015
Maulana Abul Kalam Azad University of Technology, West Bengal
2013
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 therapeutics that target the diseased conventional outflow tissue responsible for ocular hypertension in final stages human testing. The rho kinase inhibitors have proven particularly efficacious and additive therapies. Unfortunately,...
Inner retina in Alzheimer's Disease (AD) may experience neuroinflammation resulting atrophy. The objective of our study was to determine whether retinal GCIPL (ganglion cell-inner plexiform layer) or nerve fiber layer (NFL) thickness serve as noninvasive biomarkers diagnose AD. This cross-sectional case-control enrolled 15 mild cognitive impairment (MCI) patients, mild-moderate AD and 18 cognitively normal adults. NFL thicknesses on optical coherence tomography (OCT) were measured using Duke...
Abstract Müller glia, the most abundant glia of vertebrate retina, have an elaborate morphology characterized by a vertical stalk that spans retina and branches in each retinal layer. play diverse, critical roles homeostasis, which are presumably enabled their complex anatomy. However, much remains unknown, particularly mouse, about anatomical arrangement cells arbors, how these features arise development. Here we use membrane‐targeted fluorescent proteins to reveal fine structure mouse...
This paper aims toward improving background suppression from video frames by incorporating multiresolution features in Gaussian mixture model (GMM). GMM has proven its place for modeling due to better applicability and robustness compared with other popular methods literature. However, fails a number of situations such as noisy non-stationary background, slow foregrounds, illumination variation. Extensions have also been proposed increase accuracy expense increased complexity, decrease...
Purpose: To correlate ellipsoid zone (EZ) defects on spectral-domain optical coherence tomography (SD-OCT) with retinal sensitivity loss macular integrity assessment (MAIA) microperimetry in telangiectasia type 2 (MacTel). Methods: Macular SD-OCT volumes and maps were obtained during the international, multicenter, randomized phase trial of ciliary neurotrophic factor for MacTel two visits within 5 days one another. Software was developed to register MAIA scanning laser ophthalmoscopy images...
Segmentation of a medical image based on the modeling and estimation tissue intensity probability density functions via Gaussian mixture model has recently received great attention. However, distribution is unbounded symmetrical around its mean. This study presents new bounded asymmetric for analyzing both univariate multivariate data. The advantage proposed that it flexibility to fit different shapes observed data such as non-Gaussian, nonsymmetric, support Another each component ability...
The proposed work is targeted toward improving the Gaussian mixture model (GMM) for background suppression-based moving object detection. GMM has been widely used detection due to its high applicability. However, cannot properly noisy or nonstationary backgrounds and fails discriminate between foreground modes. extensions provide increased accuracy in expense of complex implementation reduced In response, this proposes two simple improvements: 1) a novel distance measure based on local...
We use semiautomated segmentation of fluorescein angiography (FA) to determine whether anti-vascular endothelial growth factor (VEGF) treatment for diabetic macular edema (DME) differentially affects microaneurysm (MA)-associated leakage, termed focal versus non-MA-associated diffuse leakage.We performed a retrospective study 29 subjects treated with at least three consecutive injections anti-VEGF agents DME (mean 4.6 injections; range, 3-10) who underwent Heidelberg FA before and after...
Patient motion artifacts are often visible in densely sampled or large wide field-of-view (FOV) retinal optical coherence tomography (OCT) volumes. A popular strategy for reducing is to capture two orthogonally oriented volumetric scans. However, due larger volume sizes, longer acquisition times, and corresponding artifacts, the registration of FOV scans remains a challenging problem. In particular, gaps data eye motion, such as saccades, can be significant their modeling becomes critical...
We present a novel multiresolution analysis based stereo matching method using curvelets and modified adaptive support weight. Multiresolution has long been applied to correspondence. However, previous methods suffer from false matches arising textureless region or repetitive textures fattening effect due area matching. In the proposed approach, we have reduced by curvelet coefficients in different scales orientations. Curvelet can uniquely represent image points increase accuracy. The is...
Mixture models for video segmentation have mainly revolved around Gaussian distributions a long time due to their simplicity and applicability. In thiswork, we proposeanovel real-time algorithm based on Student's t mixture model. Though, t-distribution has been used image by applying Expectation Maximization(EM) algorithm,the same technique cannotbe followedin videosegmentationduetoexceptional increase in computational complexity. Thus,in spiteof beinga more heavily-tailed distribution...
Gestational age estimation at time of birth is critical for determining the degree prematurity infant and administering appropriate postnatal treatment. We present a fully automated algorithm estimating gestational premature infants through smartphone lens imaging anterior capsule vasculature (ALCV). Our uses convolutional network blind image quality analyzers to segment usable regions. Then, it extracts ALCV features using residual neural architecture trains on these support vector...
Intro Anterior lens capsule vascularity (ALCV) is resorbed in the developing fetus from 27 to 35 weeks gestation. In this pilot study, we evaluated feasibility and validity of combining smartphone ophthalmoscope videos ALCV image analysis for gestational age estimation. Methods were captured longitudinally preterm neonates delivery using a PanOptic® Ophthalmoscope with an iExaminer® adapter (Welch-Allyn). video frames manually selected quantified semi-automatic analysis. A predictive model...
Development of an automatic streaming video segmentation method is crucial for many analysis applications. However, consistency temporal and scalability real-time applications are difficult to achieve. This work proposes a linear-time which scalable temporally consistent videos. A Gaussian Mixture Model (GMM) used segment each frame while recursive filtering updates the parameters GMM. hybrid methodology can uniquely propagate clusters through new frame, update variance recursively, create...
The problem of distributed estimation the probability density function (PDF) any environmental from sensor network measurement is addressed. proposed algorithm estimate local spatial parameter some as well global parameters in manner by fusing among neighboring nodes. data modeled using Gaussian mixture PDFs and an to maximizing log likelihood data. This for has been validated simulated Also real world a used function.
Motion segmentation has been a well explored research topic due to its vast application area. This work proposes real-time motion method based on 3D histogram and temporal mode selection. The distribution of video sequence consists the in foreground relatively immobile background. A provides short-term memory aforementioned distribution. selection process involves identifying most frequent values construct background thereafter. detailed analysis proposed along with an easy-to-implement...
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 motion algorithm. The provides general platform to use any MR analysis method 1) incorporate subbands information containing varying features enhanced extraction 2)...
Müller glia, the most abundant glia of vertebrate retina, have an elaborate morphology characterized by a vertical stalk that spans retina and branches in each retinal layer. play diverse, critical roles homeostasis, which are presumably enabled their complex anatomy. However, much remains unknown, particularly mouse, about anatomical arrangement cells arbors, how these features arise development. Here we use membrane-targeted fluorescent proteins to reveal fine structure mouse arbors. We...
Background suppression in video sequences has recently received great attention. While there exist many algorithms for background suppression, an important and challenging issue arising from these studies concerns that which attributes of the data should be used modelling. It is interesting difficult because no knowledge about to guide search. Also, real application length unknown frames are generated dynamically a streaming fashion arrive one at time. Thus, it impractical wait until all...