Patents - stay tuned to the technology

Inventors list

Assignees list

Classification tree browser

Top 100 Inventors

Top 100 Assignees


Utkarsh

Utkarsh Gupta, Gwalior IN

Patent application numberDescriptionPublished
20150088953METHODS, SYSTEMS AND COMPUTER-READABLE MEDIA FOR DISTRIBUTED PROBABILISTIC MATRIX FACTORIZATION - The present invention provides a method and system for distributed probabilistic matrix factorization. In accordance with a disclosed embodiment, the method may include partitioning a sparse matrix into a first set of blocks on a distributed computer cluster, whereby a dimension of each block is MB rows and NB columns. Further, the method shall include initializing a plurality of matrices including first mean matrix Ū, a first variance matrix Ũ, a first prior variance matrix Ũ03-26-2015

Utkarsh Prateek, Woburn, MA US

Patent application numberDescriptionPublished
20120047120SEARCH ENGINE OPTIMIZATION ASSISTANT - Website optimization methods and tools for optimizing visibility of a website to internet search engines are presented. A website is automatically evaluated against one or more optimization tests based on one or more search terms. A report is generated along with tools that guide and prompt the user for user input that is used by the tools to directly edit content of the website to improve the, visibility of the website to Internet search engines.02-23-2012
20140164345SEARCH ENGINE OPTIMIZATION ASSISTANT - Website optimization methods and tools for optimizing visibility of a website to internet search engines are presented. A website is automatically evaluated against one or more optimization tests based on one or more search terms. A report is generated along with tools that guide and prompt the user for user input that is used by the tools to directly edit content of the website to improve the visibility of the website to internet search engines.06-12-2014

Utkarsh Shah, Fayetteville, AR US

Patent application numberDescriptionPublished
20140154393CONJUGATED LINOLEIC ACID RICH VEGETABLE OIL PRODUCTION FROM LINOLEIC RICH OILS BY HETEROGENEOUS CATALYSIS - The invention is generally directed to conjugated linoleic acid (CLA)-rich vegetable oil production from linoleic rich oils by heterogeneous catalysis. A heterogeneous catalytic vacuum distillation process is used under high temperature conditions to isomerize linoleic acid in triacylglyceride vegetable oils to CLA to produce CLA-rich oils. After processing, the catalyst may be removed by filtration or centrifugation to obtain high quality, CLA-rich oils. The CLA-rich oils may then serve as a potent and bioactive nutraceutical and can be incorporated into various food products, such as a CLA-rich dressing, margarine or chips.06-05-2014

Utkarsh Sharma, San Leandro, CA US

Patent application numberDescriptionPublished
20130208240SEGMENTATION AND ENHANCED VISUALIZATION TECHNIQUES FOR FULL-RANGE FOURIER DOMAIN OPTICAL COHERENCE TOMOGRAPHY - Presented here are new processing techniques for optical coherence tomography (OCT) data that allow for improved visualization and use of full-range OCT images. These techniques minimize the central line artifact and the complex conjugate artifact without requiring additional system hardware or significantly increasing post-processing time. The central line artifact is minimized by normalizing each A-scan to account for ripples at the zero-delay position. The complex conjugate artifact is minimized by segmentation of a layer or layers that cross the zero-delay position, and in some embodiments by further segmentation of other surfaces based on the segmentation of the initial layer or layers. The segmentation information is then used to selectively attenuate the complex conjugate image. It may also be used for other purposes, such as dewarping.08-15-2013
20130289882INTER-FRAME COMPLEX OCT DATA ANALYSIS TECHNIQUES - Techniques for collecting and processing complex OCT data to detect localized motion contrast information with enhanced accuracy and sensitivity are presented. In a preferred embodiment, vector differences between complex OCT signals taken at the same location on the sample are used to detect blood flow in the retina. Additional embodiments involving non-linear intensity weighting of the motion contrast information, normalization of the vector difference amplitudes, and calculating the absolute value of the standard deviation of Doppler signal are described. Image processing techniques to enhance the images resulting from these motion contrast techniques are also presented.10-31-2013
20130301000SYSTEMS AND METHODS FOR FASTER OPTICAL COHERENCE TOMOGRAPHY ACQUISITION AND PROCESSING - Systems and methods for faster acquisition and processing of OCT image data are presented. In one embodiment of the present invention, an OCT system is operated in two different acquisition modes of different axial resolutions to allow for efficient collection of OCT Angiography data while also collecting high resolution OCT data. In another embodiment, a reduced subset of a collected data set is used for OCT Angiography data analysis. In another embodiment, a sweep of a swept-source laser is split into different spectral components covering different transverse locations on the sample. A further aspect includes the ability to process all or a portion of collected OCT data with one motion contrast technique before or while a second set is processed using a different motion contrast technique.11-14-2013
20130301008ANALYSIS AND VISUALIZATION OF OCT ANGIOGRAPHY DATA - Methods for analyzing and visualizing OCT angiography data are presented. In one embodiment, an automated method for identifying the foveal avascular zone in a two dimensional en face image generated from motion contrast data is presented. Several 3D visualization techniques are presented including one in which a particular vessel is selected in a motion contrast image and all connected vessels are highlighted. A further embodiment includes a stereoscopic visualization method. In addition, a variety of metrics for characterizing OCT angiography image data are described.11-14-2013
20140029015SYSTEMS AND METHODS FOR ARTIFACT SUPPRESSION BY FREQUENCY SHIFTING OR CAVITY LENGTH ADJUSTMENT - Swept source designs that eliminate or significantly reduce artifacts in optical coherence tomography are presented. One embodiment of the present invention is a source design that frequency shifts the coherence revival interference signal to a frequency larger than the A/D detection bandwidth or the post-processing bandwidth. In another embodiment, the introduced frequency shift is large enough to introduce a Doppler shift of the modes of the laser, which causes a blurring of the comb function, and thus eliminates or reduces mode hopping. In another embodiment, adjusting the cavity optical path length prior to data acquisition depending on the given optical layout configuration to reduce or eliminate coherence revival artifacts is described.01-30-2014
20140268046SYSTEMS AND METHODS FOR IMPROVED ACQUISITION OF OPHTHALMIC OPTICAL COHERENCE TOMOGRAPHY DATA - Systems and methods for improved acquisition of ophthalmic optical coherence tomography data are presented, allowing for enhanced ease of use and higher quality data and analysis functionality. Embodiments include automated triggering for detecting and initiating collection of OCT ophthalmic data, an automated technique for determining the optimal number of B-scans to be collected to create the highest quality image and optimize speckle reduction, automated review of fundus images collected with an adjunct imaging modality to guide the OCT data collection, a single scan protocol in which a large field of view is collected with HD B-scans embedded at different locations depending on automated analysis of either a fundus image or sparse OCT scan, and various scan configurations for imaging eyes with large axial depth range.09-18-2014
20140276025MULTIMODAL INTEGRATION OF OCULAR DATA ACQUISITION AND ANALYSIS - Regions-of-interest discovered from analyses of images obtained from one imaging modality can be further observed, analyzed, supplemented, and further analyzed by one or more additional imaging modalities and in an automated way. In addition, one or more pathologies identified from analyses of these regions-of-interest, and a metric of the likelihood of the presence of disease, and/or a metric of risk of disease progression can be derived therefrom.09-18-2014

Utkarsh Sharma, Dublin, CA US

Patent application numberDescriptionPublished
20150062532DATA ACQUISITION METHODS FOR REDUCED MOTION ARTIFACTS AND APPLICATIONS IN OCT ANGIOGRAPHY - Systems and methods for reducing the effects of motion on functional optical coherence tomography (OCT) imaging are described. Embodiments including post-processing and motion tracking are presented. A preferred embodiment in which functional OCT data is collected and analyzed for motion as a multiple scan unit is described. An extension of the invention to the collection of large field of view or montaged functional OCT data sets is also presented.03-05-2015
20160000320SYSTEMS AND METHODS FOR SPECTRALLY DISPERSED ILLUMINATION OPTICAL COHERENCE TOMOGRAPHY - Systems and methods are presented for acquisition and processing of spectrally dispersed illumination optical coherence tomographic data. Light from a source is distributed spectrally on the sample, and each acquisition simultaneously provides partial spectral interference information from multiple locations in the sample. Thus for a given spatial point, a single observation will be of a partial spectrum A-scan. When multiple partial spectrum A-scan observations are made at the same point by shifting the spectrum of light on to the tissue, the point can be observed by the entire broadband spectrum of the light source, thereby making it possible to create a full axial resolution A-scan.01-07-2016
20160038021SYSTEMS AND METHODS FOR ENHANCED ACCURACY IN OCT IMAGING OF THE CORNEA - Systems and methods for enhanced accuracy in optical coherence tomography imaging of the cornea are presented, including approaches for more accurate corneal surface modeling, pachymetry maps, keratometric values, and corneal power. These methods involve new scan patterns, an eye tracking mechanism for transverse motion feedback, and advanced motion correction algorithms. In one embodiment the methods comprise acquiring a first sparse set of data, using that data to create a corneal surface model, and then using the model to register a second set of denser data acquisition. This second set of data is used to create a more accurate, motion-corrected model of the cornea, from which pachymetry maps, keratometric values, and corneal power information can be generated. In addition, methods are presented for determining simulated keratometry values from optical coherence tomography data, and for better tracking and registration by using both rotation about three axes and the corneal apex.02-11-2016
20160040977METHODS OF REDUCING MOTION ARTIFACTS FOR OPTICAL COHERENCE TOMOGRAPHY ANGIOGRAPHY - A method for reducing motion artifacts in optical coherence tomography (OCT) angiography images is disclosed. The method is applied to the intensity or complex OCT data prior to applying the motion contrast analysis and involves determining sub-pixel level shifts between at least two B-scans repeated approximately at the same location and applying the sub-pixel level shifts to the B-scans to be able to correct for motion and accurately determine motion contrast signal. A preferred embodiment includes the use of 2D cross correlations to register a series of B-scans in both the axial (z-) and lateral (x-) dimensions and a convolution approach to achieve sub-pixel level frame registration.02-11-2016
20160069663SPLIT INTEGRATION MODE ACQUISITION FOR OPTIMIZED OCT IMAGING AT MULTIPLE SPEEDS - Split Integration Mode (SIM) acquisition schemes are presented that enable optical coherence tomography (OCT) imaging at multiple rates. SIM enables a system and method of operation having a first mode and a second mode, wherein the fundamental acquisition rate of the detector is the same in the two modes, but wherein the generated signals in the second mode are digitally combined prior to signal processing to create a data set with an effective acquisition rate less than the fundamental acquisition rate.03-10-2016

Utkarsh Sinha, San Jose, CA US

Patent application numberDescriptionPublished
20160114488CUSTOMER SERVICE ROBOT AND RELATED SYSTEMS AND METHODS - A robot for providing customer service within a facility includes a locomotion platform, an upper sensor for detecting objects within an upper field of view of the robot, a lower sensor for detecting objects within a lower field of view of the robot, a display and a robot computer in communication with the locomotion platform, the upper sensor and the lower sensor. The robot computer is configured to detect the presence of a customer within the facility based on information received from at least one of the upper sensor and lower sensor, and the robot computer is further configured to access one or more databases storing information associated with products available to customers within the facility and to provide customer service to the customer based on the accessed information.04-28-2016
Website © 2016 Advameg, Inc.