International Business Machines Corporation

Globalfoundries Inc.

Anna Wanda Topol

Daniel James Winarski

Stephen Leonard Schwartz

Allen Keith Bates

Nils Haustein

Dylan J Boday

Sampath Purushothaman

Bruce Kenneth Furman

Francois Pagette

Thorsten Krause

Christian Lavoie

Wayne Isami Imaino

Cornelia K Tsang

Kuan-Neng Chen

David Berman

Icko E T Iben

James Johnson

Douglas C La Tulipe, Jr

Randolph F Knarr

Frank Eliot Levine

Cyril Cabral, Jr

Sandra Guy Malhotra

Louie Arthur Dickens

Stephen Mark Rossnagel

Mary Prilotski Kusko

David Lee Rath

David Michael Fried

Albert M Young

Xiaoyan Shao

Steven Erik Steen

Michael Joseph Rooks

David C Ahlgren

Philippe M Vereecken

John Michael Hergenrother

Sharee Jane McNab

Stella Lee Taylor

Sandra Lynn Waters

Bruce Bennett Doris

Maharaj Mukherjee

Hariklia Deligianni

Craig Anthony Klein

Michael Patrick Chudzik

Eren Kursun

Edmund Juris Sprogis

Robert H Dennard

Thomas W Dyer

John Maxwell Cohn

Rajarao Jammy

Chandrasekhar Narayan

David Robert Medeiros

Matthew R Wordeman

Leathen Shi

Rama Divakaruni

Gregory Gower Freeman

Lee Curtis Randall

Joseph Francis Shepard, Jr

Ulf Troppens

Yada Zhu

David James Frank

Timothy A Johnson

Christopher M Schnabel

Yun Mou

Dylan Boday

Guy M Cohen

Keith Kwong Hon Wong

Andrew Herbert Simon

Meikei Ieong

Yun-Yu Wang

Muthumanickam Sankarapandian

Xuan Liu

Ajay Ashok Deshpande

Jae-Eun Park

Christopher D Sheraw

Wayne Alan McKinley

James Vichiconti

Steven Mark Groff

David E Cox

Sven Oehme

Icko Eric Timothy Iben

Susan K Schreitmueller

Marc Thadeus Roskow

Icko Et Iben

Zixuan Yuan

Alexander Abderrazag

Tara L Awezec

Lili Deligianni

Unsupervised cross-domain data augmentation techniques for long-text document based prediction and explanation are provided. In one aspect, a system for long-document based prediction includes: an encoder for creating embeddings of long-document texts with hierarchical sparse self-attention, and making predictions using the embeddings of the long-document texts; and a multi-source counterfactual augmentation module for generating perturbed long-document texts using unlabeled sentences from at least one external source to train the encoder. A method for long-document based prediction is also provided.

Unsupervised cross-domain data augmentation for long-document based prediction and explanation

A system, method and computer program product for archiving image, audio, and text data with metadata encapsulation in a client-server storage library is described. The server receives and holds the images, audio, or text to be archived in an image, audio or text logical partition which includes a directory of the images, audio, or text. The information is encapsulated in a metadata wrapper and stored in the library as a closed image, audio, or text file along with a closed copy of the directory. The closed image, audio, or text directory is also stored in the client. The images may be encapsulated in MXF, DICOM, Tape Archive (TAR) or GZIP formats. The storage library may have magnetic tapes, magnetic disks or optical disks as storage media.

Client-server multimedia archiving system with metadata encapsulation

The inventive concepts presented herein relate to methods of identifying molecules identification of molecules using apparatuses including: electromagnetic write-head(s); magneto-resistive read sensor(s), and processor(s). An exemplary method includes scanning one or more of a plurality of biosample tracks of a biosample substrate using an electromagnetic write-head to magnetically excite one or more molecules of interest using an alternating magnetic field. The one or more molecules of interest are disposed on one or more of the biosample tracks. A resonant response of the one or more molecules of interest is measured using a magneto-resistive sensor. The resonant response is compared to a table of known resonant responses to identify the one or more molecules of interest.

Identification of molecules based on frequency responses using electromagnetic write-heads and magneto-resistive sensors

A hypergraph is constructed based on historical shopping cart data. A node of the hypergraph corresponds to a shopping basket, and a hyperedge of the hypergraph corresponds to a unique product, the hyperedge connecting all nodes of the hypergraph representing baskets containing the unique product. A hypergraph partition algorithm identifies a cluster of shopping baskets represented in the hypergraph and determined to be similar to a given basket. Based on the cluster of shopping baskets a dual-level return prediction is performed. The dual-level return prediction includes predicting whether the given basket will be returned, and based on predicting that the given basket will be returned, predicting a probability that a product in the given basket will be returned. Based on predicting that the given basket will be returned, an ameliorative action is performed to reduce the probability.

Hypergraph structure and truncation method that reduces computer processor execution time in predicting product returns based on large scale data

Described are embodiments of an invention for detecting target antigens in a biological sample using a sample assembly. Detection may be accomplished by performing a method comprising: sweeping a head module over the sample assembly, wherein said head module includes at least one magneto-resistive read sensor configured to detect target antigens via nanoparticles within the sample assembly; and detecting at least one particular antigen among the target antigens. Preferably, detecting the target antigens via the nanoparticles is based at least in part on detecting unique magnetic properties of particular nanoparticles specifically associated with different types of the target antigens. Detection using a magnetic read/write head in the sample assembly facilitates automation of sample detection with high speed and fidelity. Corresponding systems are also disclosed.

Trenched sample assembly for detection of analytes with electromagnetic read-write heads

Described are embodiments to calibrate read sensors, which in turn may ensure that the equipment utilized to detect antigens is reliable and accurate. If it is determined that a read sensor is degraded a method of calibrating a read sensor of a read head may be used.

Calibrating read sensors of electromagnetic read-write heads

Described are embodiments of an invention for a sample assembly with an electrical conductor for detection of the antigens by electromagnetic read heads. In one embodiment, a sample assembly includes: a substrate; one or more base layers above the substrate; an outer layer above the substrate; a plurality of sample trenches formed in the outer layer, each sample trench being characterized by an upper surface, a bottom surface, and a longitudinal axis; an electrical conductor disposed in the substrate, the electrical conductor being configured to generate an electromagnetic field in proximity to the plurality of sample trenches to enhance nanoparticle movement toward the bottom surface of the plurality of sample trenches; and at least one alignment trench formed above the substrate, each alignment trench having a longitudinal axis substantially parallel to a longitudinal axis of at least one of the sample trenches.

Sample assembly with an electromagnetic field to accelerate the bonding of target antigens and nanoparticles

The inventive concepts presented herein relate to methods of identifying molecules identification of molecules using apparatuses including: electromagnetic write-head(s); magneto-resistive read sensor(s), and processor(s). An exemplary method includes magnetically exciting a molecule to be identified using an alternating magnetic field generated by an electromagnetic write-head, measuring a resonant response of the molecule to be identified using a magneto-resistive read sensor; and comparing, using a processor, the resonant response of the molecule to be identified with a table of known resonant responses to identify a chemical composition of the molecule to be identified. The molecule to be identified may optionally be disposed on a biosample substrate which comprises, or is coupled to, a plurality of servo-alignment marks; and the plurality of servo-alignment marks are configured to facilitate alignment of the electromagnetic write-head with the biosample tracks of the biosample substrate.

Embodiments of the disclosure relate to a biosample plate that includes a memory component for storing biosample identification and analysis data, and a wireless communication interface for transferring the data to and from the biosample plate. In one embodiment, the biosample plate comprises a base for receiving a biosample, a memory component coupled to the base for storing identification and analysis information related to the biosample, and a wireless communication interface coupled to the memory component for transferring the information to and from the memory component. The wireless communication interface may include an electromagnetic device.

Biosample plate with data storage and wireless communications

A sample assembly includes an outer layer with at least one sample trench. The sample trench includes a first set of antibodies that are bonded on a first surface of a base layer. Target antigens are bonded with the first set of antibodies, and a second set of antibodies are bonded to the target antigens. Further, the sample trench includes nanoparticles that are bonded to the second set of antibodies.

Embodiments of the disclosure relate to a biosample plate that includes a memory component for storing biosample identification and analysis data, and a wireless communication interface for transferring the data to and from the biosample plate. In one embodiment, the biosample plate comprises a base for receiving a biosample, a memory component coupled to the base for storing identification and analysis information related to the biosample, and a wireless communication interface coupled to the memory component for transferring the information to and from the memory component. The wireless communication interface may include an optical device.

Described are embodiments of an invention for a sample assembly with an electrical conductor for detection of the antigens by electromagnetic read heads. In one embodiment, a sample assembly includes: a substrate; one or more base layers above the substrate; an outer layer above the one or more base layers; a plurality of sample trenches formed in the outer layer and/or the one or more base layers, each sample trench being characterized by an upper surface, a bottom surface, and a longitudinal axis; an electrical conductor disposed in the substrate, the electrical conductor being configured to generate an electromagnetic field in proximity to the plurality of sample trenches to enhance nanoparticle movement toward the bottom surface of the plurality of sample trenches; and at least one alignment trench formed above the substrate, each alignment trench having a longitudinal axis substantially parallel to a longitudinal axis of at least one of the sample trenches.

Embodiments of the disclosure relate to methods of storing and using biosamples with a cartridge that includes slots for storing biosample capillary tubes. The methods include providing access to a holder inside an enclosure of the cartridge, the enclosure having a same form factor as a data tape cartridge used in an automated tape library. The holder is configured to receive a plurality of capillary tubes in the holder, the capillary tubes including one or more biosamples. Exemplary methods may also include receiving capillary tubes in the cartridge, withdrawing capillary tubes from the cartridge, scanning and/or analyzing the capillary tubes and/or biosamples. Exemplary methods may additionally or alternatively include retrieving the cartridge from a storage slot of the automated tape library, loading the cartridge onto a drive of the automated tape library, and/or receiving and/or exporting a cartridge to/from the automated tape library via a mail slot.

Cartridge for storing biosample capillary tubes and use in automated data storage systems

Growing community of inventors

Jefferson Valley, NY, United States of America

Anna Wanda Topol