Siemens Healthcare Gmbh

Siemens Healthcare Diagnostics Gmbh

Siemens Healthineers Ag

Other

Siemens Aktiengesellschaft

Siemens Corporation

Vivek Kumar Singh

Terrence Chen

Yao-Jen Chang

Kai Ma

Bogdan Georgescu

Birgi Tamersoy

Dorin Comaniciu

Ankur Kapoor

Yefeng Zheng

Andreas Wimmer

Hien Nguyen

Ali Kamen

Benjamin S Pollack

Puneet Sharma

Jiangping Wang

Frank Sauer

Stefan Kluckner

Andreas Krauss

Klaus J Kirchberg

Brian Teixeira

Venkatesh NarasimhaMurthy

Ruhan Sa

Shaohua Kevin Zhou

David D Liu

Ralf Nanke

Rayal Raj Prasad Nalam Venkat

Tommaso Mansi

Sasa Grbic

Li Zhang

Rui Liao

Shun P Miao

Mehmet Akif Gulsun

Michael Wels

Daguang Xu

Julian Krebs

Florin Cristian Ghesu

Pierre De Tournemire

Shanhui Sun

Boris Mailhe

Grzegorz Soza

Susanne Dornberger

Susanne Oepping

Philipp Hölzer

Yifan Wu

Mariappan S Nadar

Ziyan Wu

Tiziano Passerini

Matthias John

Daphne Yu

Thomas J O'Donnell

Eli Gibson

Mingqing Chen

Jan Kretschmer

Nathan S Lay

Thomas Re

Mark Edwards

Max Schöbinger

John Paulus, Jr

Eva Eibenberger

Diana Ioana Stoian

Florin-Cristian Ghesu

Alexandru Turcea

Rui Wang

Oliver Lehmann

Sarfaraz Hussein

Amitkumar Bhupendrakumar Shah

Andreas Prokein

Subhabrata Bhattacharya

Tobias Betz

Santosh Pai

Liheng Zhang

Elena Balashova

Klaus Kirchberg

Andreas Krauß

Zhuokai Zhao

Jianping Wang

Lequan Yu

Venkatesh N Murthy

A method of training a model of a diagnostic apparatus includes providing one or more first tube assemblies of a first type and one or more second tube assemblies of a second type in a diagnostic apparatus; capturing one or more first images of at least a portion of each of the one or more first tube assemblies and the second tube assemblies using the imaging device. Training the model includes identifying tube assemblies of the first type and tube assemblies of the second type based on the one or more first images and the one or more second images. Tubes assemblies of the first type are grouped into a first group and tube assemblies of the second type are grouped into a second group. Other methods and apparatus are disclosed.

Apparatus and methods of training models of diagnostic analyzers

A method of synthesizing an image of a tube assembly includes capturing an image of the tube assembly, wherein the capturing generates a captured image. The captured image is decomposed into a plurality of features in latent space using a trained image decomposition model. One or more of the features in the latent space is manipulated into one or more manipulated features. A synthesized tube assembly image is generated with at least one of the manipulated features using a trained image composition model. Other methods and systems are disclosed.

Diagnostic laboratory systems and methods of imaging tube assemblies

A method of identifying a tube type. The method includes capturing one or more pixelated images of a cap affixed to a tube; identifying a color of one or more pixels of the pixilated image of the cap; identifying one or more gradients of a dimension of the cap; and identifying the tube type based at least on: the color of the one or more pixels, and the one or more gradients of a dimension of the cap. Apparatus adapted to carry out the method are disclosed as are other aspects.

Apparatus and methods of identifying tube assembly type

A method of characterizing a specimen and specimen container to be analyzed in an automated diagnostic analysis system. The method can provide a segmentation determination and/or an HILN determination (hemolysis, icterus, lipemia, or normal) of the specimen while protecting patient information. The method includes capturing an image of a specimen container via an image capture device, identifying a label affixed to the specimen container in the captured image via an anonymization network, and editing the captured image via the anonymization network to mask some or all information present in the label so that it is removed from the captured image. Quality check modules and systems configured to carry out the method are also described, as are other aspects.

Methods and apparatus for protecting patient information during characterization of a specimen in an automated diagnostic analysis system

Anomalies in images are detected. A generative network and/or an autoencoder ('G/A-Network'), a Siamese network, a first training-dataset of normal images and a second training-dataset of abnormal images are provided. The G/A-network is trained to produce latent data from input images and output images from the latent data, wherein the training is performed with images of the first training-dataset, wherein a loss function is used for training at least at the beginning of training, and the loss function enhances the similarity of the input images and respective output images. The Siamese network is trained to generate similarity measures between input images and respective output images, wherein the training is performed with images of the first training-dataset and the second training-dataset in that images of both training-datasets are used as input images for the G/A-network and output images of the G/A-network are compared with their respective input images by the Siamese network.

System and method for detecting anomalies in images

A method of characterizing a specimen to be analyzed in an automated diagnostic analysis system provides an HILN classification (hemolysis, icterus, lipemia, normal) of the specimen along with a basis for that determination. The method includes assigning a hash code to each training image of a sample specimen used in the characterization training process. In response to an HILN determination for a test specimen, the method can retrieve via the hash code one or more of the closest matching training images upon which the HILN classification is based. The one or more of the closest matching training images can be displayed alongside of the one or more images of the test specimen. Quality check modules and systems configured to carry out the method are also described, as are other aspects.

Methods and apparatus for hashing and retrieval of training images used in HILN determinations of specimens in automated diagnostic analysis systems

A stream of virtual topograms, in particular live virtual topograms, is predicted. Sets of surface data of an outer surface of a subject are continuously received. Based on each received set of surface data a (live) virtual topogram is continuously generated by a trained machine learning algorithm (MLA). Thereto, a representation of body landmarks is updated based on each received set of surface data by a trained body marker detector (BMD), of the trained MLA, and the (live) virtual topogram is predicted based on the updated spatial marker map and on the corresponding set of surface data by a trained topogram generator (TG) of the trained MLA.

Method and systems for predicting a stream of virtual topograms

A computer-implemented method for detecting a failure of a device connected to a sensor is disclosed. The method includes a machine learning model receiving a trace signal from the sensor indicating a status of the device, the machine learning model encoding the trace signal into a plurality of vector representations, and the machine learning model determining whether the trace signal is valid or invalid based on the plurality of vector representations.

Automatic sensor trace validation using machine learning

Systems and methods for generating a probabilistic tree of vessels are provided. An input medical image of vessels of a patient is received. Anatomical landmarks are identified in the input medical image. A centerline of the vessels in the input medical image is determined based on the anatomical landmarks. A probabilistic tree of the vessels is generated based on a probability of fit of the anatomical landmarks and the centerline of the vessels. The probabilistic tree of the vessels is output.

Probabilistic tree tracing and large vessel occlusion detection in medical imaging

Methods and systems for image registration using an intelligent artificial agent are disclosed. In an intelligent artificial agent based registration method, a current state observation of an artificial agent is determined based on the medical images to be registered and current transformation parameters. Action-values are calculated for a plurality of actions available to the artificial agent based on the current state observation using a machine learning based model, such as a trained deep neural network (DNN). The actions correspond to predetermined adjustments of the transformation parameters. An action having a highest action-value is selected from the plurality of actions and the transformation parameters are adjusted by the predetermined adjustment corresponding to the selected action. The determining, calculating, and selecting steps are repeated for a plurality of iterations, and the medical images are registered using final transformation parameters resulting from the plurality of iterations.

Method and system for image registration using an intelligent artificial agent

CT scan parameters for performing a CT scan of an anatomical target region of a patient are determined and/or adjusted. An initial set of the CT scan parameters for starting to perform the CT scan is determined based on an initial set of attenuation curves associated with the anatomical target region of the patient. The initial set of attenuation curves are determined based on optical imaging data depicting the patient.

Determining CT scan parameters based on machine learning

A 3D shape is reconstructed from a topogram. A generative network is machine trained. The generative network includes a topogram encoder for inputting the topogram and a decoder to output the 3D shape from the output of the encoder. For training, one or more other encoders are included, such as for input of a mask and/or input of a 3D shape as a regularlizer. The topogram encoder and decoder are trained with the other encoder or encoders outputting to the decoder. For application, the topogram encoder and decoder as trained, with or without the encoder for the mask and without the encoder for the 3D shape, are used to estimate the 3D shape for a patient from input of the topogram for that patient.

Three-dimensional shape reconstruction from a topogram in medical imaging

Machine learning is used to train a network to estimate a three-dimensional (3D) body surface and body regions of a patient from surface images of the patient. The estimated 3D body surface of the patient is used to determine an isocenter of the patient. The estimated body regions are used to generate heatmaps representing visible body region boundaries and unseen body region boundaries of the patient. The estimation of 3D body surfaces, the determined patient isocenter, and the estimated body region boundaries may assist in planning a medical scan, including automatic patient positioning.

Dense body marker estimation from camera data for patient positioning in medical imaging

A method of characterizing a serum or plasma portion of a specimen in a specimen container provides an HILN (hemolysis, icterus, lipemia, normal) determination. Pixel data of an input image of the specimen container is processed by a classification network to identify whether the specimen contains plasma or serum. Pixel data representing a plasma sample are forwarded to a segmentation/classification/regression network trained with plasma samples for HILN determination. Pixel data representing a serum sample are forwarded to a transformation network, wherein the serum sample pixel data is transformed into pixel data that matches pixel data of a corresponding previously-collected plasma sample by changing sample color, contrast, intensity, and/or brightness. The transformed serum sample pixel data are forwarded to the segmentation/classification/regression network for HILN determination. Quality check modules and testing apparatus configured to carry out the method are also described, as are other aspects.

Methods and apparatus for HILN determination with a deep adaptation network for both serum and plasma samples

For training a machine learning system for representing a patient body a plurality of stored medical imaging data sets each representing at least a part of a respective patient are obtained. A first one of the plurality of stored medical imaging data sets represents a different part of the patient body than a second one of the plurality of stored medical imaging data sets. A plurality of landmarks in the stored medical imaging data sets are estimated, and each of the stored medical imaging data sets are aligned to a predefined pose using the plurality of landmarks. A plurality of points in the aligned medical imaging data sets are sampled, and the machine learning system is trained based on at least the plurality of points. The learned parameters of the machine learning system are then stored and used in a method for inferring a body representation.

Body representations

Growing community of inventors

Princeton, NJ, United States of America

Vivek Kumar Singh