Koninklijke Philips Corporation N.v.

Case Western Reserve University

Other

Covera Health, Inc.

Daniel Robert Elgort

Jeffrey L Duerk

Lucian Remus Albu

Jonathan S Lewin

Eddy Y Wong

Claudia M Hillenbrand

Ron Vianu

Benjamin L Odry

Richard J Herzog

Benjamin Sellman Suutari

Gregory Allen Dubbin

Raymond C Chan

Robert Manzke

Ladislav Jankovic

Yugang Jia

Reza Sharifi Sedeh

Khalid Shahzad

Chris Flask

Satyen Mukherjee

Christopher A Flask

Brian M Dale

Bryce Eakin

Robert Allen Epstein

James Robert Browning

JinHyeong Park

Grant Langseth

Shanon V Levenherz

Mattia Ciollaro

Greg Dubbin

Scott T Denegre

Elizabeth Sweeney

Andre Perunicic

Jefferson Chen

W Nathaniel Brown

Ushanandini Raghavan

Tarmo Henrik Aijo

Xiaojin Dong

Irwin Keller

Ken Pin Wang

Murray Becker

John Peloquin

Scott Schwartz

Nathaniel W Brown

A risk-adjusted assessment of a target facility's quality measures (e.g. mortality rate, length of stay, readmission rate, complications rate, etc.) is determined with respect to the quality measures of a broader population base. Patient cohorts are identified corresponding to particular ailments or treatments, and the target facility's risk-adjusted quality measures are determined for each cohort. When a particular quality measure for a target cohort indicates poor performance, factors that are determined to be relevant to the patients' outcomes are identified and used to create a control group of patients in the broader population who exhibit similar factors but had better outcomes than the patients of the target cohort. The care process (treatments, medications, interventions, etc.) that each of the target patients received is compared to the care process that each of the control patients received, to identify potential root-causes of the poorer performance.

Automated controlled-case studies and root-cause analysis for hospital quality improvement

A natural language understanding method begins with a radiological report text containing clinical findings. Errors in the text are corrected by analyzing character-level optical transformation costs weighted by a frequency analysis over a corpus corresponding to the report text. For each word within the report text, a word embedding is obtained, character-level embeddings are determined, and the word and character-level embeddings are concatenated to a neural network which generates a plurality of NER tagged spans for the report text. A set of linked relationships are calculated for the NER tagged spans by generating masked text sequences based on the report text and determined pairs of potentially linked NER spans. A dense adjacency matrix is calculated based on attention weights obtained from providing the one or more masked text sequences to a Transformer deep learning network, and graph convolutions are then performed over the calculated dense adjacency matrix.

Computer-implemented natural language understanding of medical reports

In an embodiment, a computer-implemented process comprises accessing a plurality of digitally stored, unstructured medical diagnostic data; digitally displaying a first subset of the medical diagnostic data, the first subset of the medical diagnostic data including at least a first set of diagnostic reports, using a computer display device, concurrently with digitally displaying one or more quality control checklists that are specific to a medical discipline represented in the first set of diagnostic reports; receiving digital input specifying one or more errors in the first set of diagnostic reports and digitally storing the digital input in association with the first subset of medical diagnostic data; training a hierarchical Bayesian machine learning model using the digital input and the first subset of medical diagnostic data; evaluating the hierarchical Bayesian machine learning model, after training, for a second subset of the medical diagnostic data, the second subset being different from the first subset, to result in outputting one or more provider error rate data; applying a grading algorithm to the one or more provider error rate data to yield one or more output provider quality score values.

Computer-implemented detection and statistical analysis of errors by healthcare providers

For training data pairs comprising training text (a radiological report) and training images (radiological images associated with the radiological report), a first encoder network determines word embeddings for the training text. A concept is generated from the operation of layers of the first encoder network, which is regularized by a first loss between the generated concept and a labeled concept for the training text. A second encoder network determines features for the training image. A heatmap is generated from the operation of layers of the second encoder network, which is regularized by a second loss between the generated heatmap and a labeled heatmap for the training image. A categorical cross entropy loss is calculated between a diagnostic quality category (classified by an error encoder) and a labeled diagnostic quality category for the training data pair. A total loss function comprising the first, second, and categorical cross entropy losses is minimized.

Computer-implemented machine learning for detection and statistical analysis of errors by healthcare providers

A database merger method () merges two or more anonymized healthcare databases (X, Y). Each anonymized healthcare database has personally identifying information anonymized including having medical care units replaced by medical care unit placeholders. In the database merger method, statistical patient feature distributions are computed for medical care unit placeholders in the anonymized healthcare databases. Medical care unit placeholders in different anonymized healthcare databases are matched by matching corresponding statistical patient feature distributions for the respective medical care unit placeholders. Patients in different anonymized healthcare databases are matched. The patient matching is performed within matched pairs of medical care unit placeholders to improve computational efficiency. A databases alignment table () is generated identifying the matched patients, or a merged anonymized healthcare database is generated that merges patient features in the different anonymized healthcare databases for each matched patient into a single patient entry.

Hospital matching of de-identified healthcare databases without obvious quasi-identifiers

Systems, methodologies, media, and other embodiments associated with automatically adapting MRI controlling parameters are described. One exemplary method embodiment includes configuring an MRI apparatus to acquire MR signal data using a non-rectilinear trajectory. The example method may also include acquiring MR signals, transforming the MR signals into image data, and selectively adapting the MRI controlling parameters based, at least in part, on information associated with the MR signals.

Adaptive imaging parameters with MRI

A new and improved method for tracking and/or spatial localization of an invasive device in Magnetic Resonance Imaging (MRI) is provided. The invention includes providing an invasive device including a marker having a chemically shifted signal source with a resonant frequency different from the chemical species of the subject to be imaged, applying a pulse sequence, detecting the resulting RF magnetic resonance signals, and determining the 3D coordinates of the marker. The invention also includes generating scan planes and reconstructing an image from the detected signals to generate an image having the marker contrasted from the subject. The invasive device includes a marker having a chemically shifted signal source which has a resonant frequency different from the chemical species of the subject to be imaged for use in tracking the device during imaging.

Chemical shift markers for improved wireless fiducial marker tracking

In magnetic resonance imaging (MRI), selected magnetic dipoles in a subject are aligned with a main magnetic field for later manipulation, and signals received after such manipulations are used to create image representations of the subject. One drawback is that even powerful magnetic fields can only align a very small percentage of dipoles in the region of the field. Electromagnetic radiation endowed with orbital angular momentum (OAM) aligns dipoles along the direction of travel of the radiation, but at a much higher percentage; as high as 100% of the dipoles in the region can be aligned. Resultantly, resonance signals emanating from the region are several orders of magnitude stronger than signals emanated using traditional MRI techniques. All electromagnetic radiation, including visible light can be endowed with OAM and used to hyperpolarize a region of interest.

Magnetic resonance imaging hyperpolarization of liquids or solids by light with orbital angular momentum

A magnetic resonance examination system comprises an RF-system for inducing resonance in polarized dipoles and receiving magnetic resonance signals from an object to be examined. A thermometry module dervies a temperature distribution of the object to be examined from the magnetic resonance signals. The magnetic resonance examination system further comprises a photonic-based hyperpolarization device with a photonic source for emitting electromagnetic radiation, a moder converter, such as a phase hologram to impart orbital angular momentum to the electromagnetic radiation and via spatial filter to select from the phase hologram a diffracted photonic beam endowed with orbital angular momentum for polarizing the dipoles via transferred orbital angular momentum.

MRI thermometry combined with hyperpolarisation device using photons with orbital angular momentum

A tissue ablation device employs one or more energy emitters () and one or more photoacoustic sensors () in a cooperative arrangement for applying a tissue ablation therapy to a tissue (). In operation, the energy emitters () emit a tissue ablation beam (TA) into a target portion of the tissue () to form a lesion () therein, and alternatively or concurrently emit a photoexcitation beam (PE) into the target portion of the tissue () to excite a photoacoustic response from the tissue (). The photoacoustic sensor(s) () sense the photoacoustic response of the tissue ().

Tissue ablation device with photoacoustic lesion formation feedback

One or more light beam endowed with photonic orbital angular momentum generating devices are mounted at preselected locations on an insertable instrument to hyperpolarize nuclear magnetic dipoles in a region of interest. The hyperpolarized nuclear magnetic dipoles are caused to resonate, generating magnetic resonance signals. A controller controls gradient coils to induce a magnetic field gradient across the region of interest, such that the frequency of the resonance signals is indicative of spatial positions. A frequency-to-position decoder converts the resonance signal frequencies into spatial positions. A video processor combines the spatial positions and a portion of a diagnostic image from a diagnostic image memory into a combined image which depicts the location of the region of interest or a portion of the instrument marked on the diagnostic image and displays the combined image on a monitor.

Active device tracking using light with orbital angular momentum to hyperpolarized MRI

A device capable of producing a high resolution chemical analysis of a sample, such as fluid, is based upon nuclear magnetic resonance (NMR) spectroscopy. The nuclear magnetic polarizations of the sample are generated by sequentially illuminating the sample with a focused beam of light carrying angular orbital angular momentum (OAM) and possibly momentum (spin). Unlike in a typical NMR used for magnetic nuclear resonance imaging (MRI) or spectroscopy, the present device does not make use of a strong magnet.

Nuclear magnetic resonance spectroscopy using light with orbital angular momentum

A localization system to localize an interventional instrument in the body of a patient. The localization system includes an electromagnetic wave source which splits an electromagnetic signal into components propagating along a probe path and a reference path, respectively. The probe path includes a signal outlet for emitting the signal at the point to be located and at least one detector for picking up the emitted signal. A correlator is used to determine the correlation between the signal components that propagated along the probe path and the reference path respectively. Knowing the length of the reference path, the unknown distance between the signal outlet and the detector in the probe path can be estimated based on the correlation information.

Electromagnetic localization system

Systems and methods display and manage wireless electronic menus that are capable of connecting to a computer terminal, a server and/or a database via one or more communication networks. The systems are capable of managing wireless electronic menus within a single site establishment or within multi-site establishments The menus have one or more sensors for determining if the menus are in motion, idle, in use not in use or docked to a charging base to define a menu status for each menu. The terminal updates the menu content of the menus based on the menu status of each menu. A proximity sensor determines a location of the menus to define a configuration of the menus with respect a table or a table configuration of tables on a floor plan. Sensors, a user input and/or the computer terminal associate the menus with a menu grouping or with the table in close proximity of the menus so that common menu content may be display by each menu. The menus display menu content for available goods and/or services offered by a provider in one or more languages as selected by users of the menus. Usage of menus is accurately tracked by the proximity sensor and the sensors of the menus to eliminate false counts and to correct for errors generated by multiple uses of the menu by multiple users. The menus provide lists of goods and/or services in a format which is capable of being sorted and/or filtered based on the information and/or multimedia data indicative of the goods and/or services.

System and method for displaying and managing electronic menus

A system for automatically adapting image acquisition parameters based on imaging and/or device tracking feedback is provided. An example system includes a subsystem for acquiring images (e.g., MR) of an object and tracking data for a device (e.g. catheter) inserted into the object and controllably moveable within the object. The system also includes an image processor for processing the images and a device tracking logic for computing device parameters (e.g., speed, direction of travel, rate of speed change, position, position relative to a landmark, device orientation). Based on the images and device parameter computations, a parameter control and adjustment logic can automatically update one or more image acquisition parameters that control the image acquisition subsystem.

System and method for adjusting image parameters based on device tracking

Growing community of inventors

New York, NY, United States of America

Daniel Robert Elgort