Siemens Healthcare Gmbh

Siemens Aktiengesellschaft

Samsung Electronics Co., Ltd.

Siemens Corporation

The Johns Hopkins University

Raychem Limited

Siemens Healthcare Diagnostics Gmbh

David D Liu

Shaohua Kevin Zhou

Dorin Comaniciu

Bogdan Georgescu

Yefeng Zheng

Daguang Xu

Atilla Peter Kiraly

Youngjun Yoo

Chenchi Luo

Nathan S Lay

Christian Tietjen

Dijia Wu

Tommaso Mansi

Vivek Kumar Singh

Bernhard Geiger

Michael Suehling

Grzegorz Soza

Hien Nguyen

Li Zhang

Michael Oliver Polley

Zhoubing Xu

Noha Youssry El-Zehiry

Jan Kretschmer

S Kevin Zhou

Ali Kamen

Hamid Rahim Sheikh

Berthold Kiefer

Adrian Barbu

Dominik Bernhardt

Chris Schwemmer

Mingqing Chen

Christopher Vincent Alvino

Gunnar Krüger

Kaimo Lin

Sascha Seifert

George Q Chen

Vincent Ordy

Philipp Wolber

Terrence Chen

Puneet Sharma

Jan Boese

Shanhui Sun

Benjamin S Pollack

Boris Mailhe

Manish Goel

Sasa Grbic

Benjamin L Odry

Yao-Jen Chang

Carol L Novak

Ankur Kapoor

Daphne Yu

Klaus Engel

David J Chazan

Viorel Mihalef

Seok-Jun Lee

Michael Kelm

Wen Wu

Michael Wels

Andreas Wimmer

Jingdan Zhang

Chi-Ming Chan

Robert Grimm

Yingmao Li

Andre De Oliveira

Thomas Pheiffer

Julian Krebs

Martin Huber

Martin Kramer

Peter Gall

Michael Sühling

Li Pan

Gyeongmin Choe

Ying Chi

Maria Jimena Costa

Dong Yang

Pradeep Barma

Max Schöbinger

Tae Soo Kim

Tao Xiong

Alexander Cavallaro

Hyejung Kim

Nora Hünemohr

Johannes Feulner

He Zhang

Pavlo Dyban

Sebastian Seifert

Yaron Anavi

Hao Xu

Paul Swoboda

Sowmya Ramakrishnan

Xun (Jason) Xu

Alexander Weiss

Ihab Kamel

An electronic device, method, and computer readable medium for an invertible wavelet layer for neural networks are provided. The electronic device includes a memory and at least one processor coupled to the memory. The at least one processor is configured to receive an input to a neural network, apply a wavelet transform to the input at a wavelet layer of the neural network, and generate a plurality of subbands of the input as a result of the wavelet transform.

System and method for invertible wavelet layer for neural networks

An electronic device includes a display, a camera, and a processing device. The processing device is configured to determine whether (i) a user's face or eyes are within the camera's field of view or (ii) a gaze of the user is directed towards the display. The processing device is also configured, in response to determining that (i) the user's face or eyes are not within the camera's field of view or (ii) the gaze of the user is not directed towards the display, to modify a user interface button presented on the display. The user interface button may represent a shutter button configured to cause the camera or another camera of the electronic device to capture one or more images.

System and method for AI enhanced shutter button user interface

An electronic device, method, and computer readable medium for 3D association of detected objects are provided. The electronic device includes a memory and at least one processor coupled to the memory. The at least one processor configured to convolve an input to a neural network with a basis kernel to generate a convolution result, scale the convolution result by a scalar to create a scaled convolution result, and combine the scaled convolution result with one or more of a plurality of scaled convolution results to generate an output feature map.

System and method for convolutional layer structure for neural networks

A method of visualizing spinal nerves includes receiving a 3D image volume depicting a spinal cord and a plurality of spinal nerves. For each spinal nerve, a 2D spinal nerve image is generated by defining a surface within the 3D volume comprising the spinal nerve. The surface is curved such that it passes through the spinal cord while encompassing the spinal nerve. Then, the 2D spinal nerve images are generated based on voxels on the surface included in the 3D volume. A visualization of the 2D spinal images is presented in a graphical user interface that allows each 2D spinal image to be viewed simultaneously.

System and method for next-generation MRI spine evaluation

An electronic device, a method, and computer readable medium for operating an electronic device are disclosed. The method includes receiving data about a state of the electronic device from one or more sensors of the electronic device. The method also includes determining whether to modify a user interface button displayed on a display of the electronic device based on the received state data and parameters of a neural network. The method further includes modifying display of the user interface button on the display of the electronic device based on determining to modify. The method additionally includes providing, to the neural network, feedback data indicating whether the user interface button was triggered within a predetermined time period after modifying the user interface button.

A method includes obtaining a first image of a scene using a first image sensor of an electronic device and a second image of the scene using a second image sensor of the electronic device. The method also includes generating a first feature map from the first image and a second feature map from the second image. The method further includes generating a third feature map based on the first feature map, the second feature map, and an asymmetric search window. The method additionally includes generating a depth map by restoring spatial resolution to the third feature map.

Asymmetric normalized correlation layer for deep neural network feature matching

A method to map a plurality of feature maps of a neural network onto a memory hierarchy includes mapping a first feature map of the plurality of feature maps to a memory in a memory hierarchy having available memory space and providing quickest access to the first feature map. The method also includes, when the first feature map expires, removing the first feature map from the memory used to store the first feature map.

Dynamic memory mapping for neural networks

Methods and systems for compressed domain progressive application of computer vision techniques. A method for decoding video data includes receiving a video stream that is encoded for multi-stage decoding. The method includes partially decoding the video stream by performing one or more stages of the multi-stage decoding. The method includes determining whether a decision for a computer vision system can be identified based on the partially decoded video stream. Additionally, the method includes generating the decision for the computer vision system based on decoding of the video stream. A system for encoding video data includes a processor configured to receive the video data from a camera, encode the video data received from the camera into a video stream for consumption by a computer vision system, and include metadata with the encoded video stream to indicate whether a decision for the computer vision system can be identified from the metadata.

Progressive compressed domain computer vision and deep learning systems

A method of capturing a photo using a mobile device is described. The method comprises detecting a signature motion of the mobile device; capturing video using a camera of the mobile device upon detecting the signature motion; and automatically identifying a start frame and an end frame of the captured images using neural networks. A mobile device for capturing a photo is also described.

Capturing a photo using a signature motion of a mobile device

A method and apparatus for automatically performing medical image analysis tasks using deep image-to-image network (DI2IN) learning. An input medical image of a patient is received. An output image that provides a result of a target medical image analysis task on the input medical image is automatically generated using a trained deep image-to-image network (DI2IN). The trained DI2IN uses a conditional random field (CRF) energy function to estimate the output image based on the input medical image and uses a trained deep learning network to model unary and pairwise terms of the CRF energy function. The DI2IN may be trained on an image with multiple resolutions. The input image may be split into multiple parts and a separate DI2IN may be trained for each part. Furthermore, the multi-scale and multi-part schemes can be combined to train a multi-scale multi-part DI2IN.

Deep image-to-image network learning for medical image analysis

To assist a physician in diagnosis of trauma involving abdominal pain, scan data representing the patient is partitioned by organ and/or region. Separate machine-learnt classifiers are provided for each organ and/or region. The classifiers are trained to indicate a likelihood of cause of the pain. By outputting results from the collection of organ and/or regions specific classifiers, the likeliest causes and associated organs and/or regions may be used by the physician to speed, confirm, or guide diagnosis of the source of abdominal pain.

Source of abdominal pain identification in medical imaging

A computer-implemented method for deriving biopsy results in a non-invasive manner includes acquiring a plurality of training data items. Each training data item comprises non-invasive patient data and one or more biopsy derived scores associated with an individual. The method further includes extracting a plurality of features from the non-invasive patient data based on the one or more biopsy derived scores and training a predictive model to generate a predicted biopsy score based on the plurality of features and the one or more biopsy derived scores.

Virtual biopsy techniques for analyzing diseases

Tissue is characterized using machine-learnt classification. The prognosis, diagnosis or evidence in the form of a similar case is found by machine-learnt classification from features extracted from frames of medical scan data. The texture features for tissue characterization may be learned using deep learning. Using the features, therapy response is predicted from magnetic resonance functional measures before and after treatment in one example. Using the machine-learnt classification, the number of measures after treatment may be reduced as compared to RECIST for predicting the outcome of the treatment, allowing earlier termination or alteration of the therapy.

Tissue characterization based on machine learning in medical imaging

For liver modeling from medical scan data, multiple modalities of imaging are used. By using multiple modalities of imaging in combination with generative modeling, a more comprehensive and informed assessment may be performed. The generative modeling may allow feedback of effects of proposed therapy on function of the liver. This feedback is used to update the liver function information based on the imaging. Based on the computerized modeling with information from various imaging modes, an output based on more comprehensive information and patient personalized modeling and feedback may be provided to assist the physician.

Liver disease assessment in medical imaging

Embodiments can provide a method for atlas-based contouring, comprising constructing a relevant atlas database; selecting one or more optimal atlases from the relevant atlas database; propagating one or more atlases; fusing the one or more atlases; and assessing the quality of one or more propagated contours.

Atlas-based contouring of organs at risk for radiation therapy

Growing community of inventors

Mountain View, CA, United States of America

David D Liu