Toshiba Medical Systems Corporation

Kabushiki Kaisha Toshiba

Canon Medical Systems Corporation

Other

Barco N.v.

Meta System S.p.a.

Vysis, Inc.

Toshiba Medical Visualization Systems Europe, Limited

Voxar Limited

Barcoview Mis Edinburgh

Metasystems Hard & Software Gmbh

Ian Poole

Jim Piper

Costas Plakas

Sean Daniel Murphy

Akinola Akinyemi

Mohammad Dabbah

Alison O'Neil

Aneta Lisowska

Pavlos Papageorgiou

Keith Goatman

Thomas Lörch

James R Piper

Brian Mohr

Andrew John Bissell

Colin John Roberts

Erin Beveridge

Jian Song

Paul Norman

Matthew Daykin

Ruixuan Wang

Jillian Steel

Thomas Richard Lorch

James Richard Piper

An apparatus includes processing circuitry configured to: obtain first trained parameters for a model, wherein the first trained parameters have been generated by training the model using data from a first data cohort; obtain second trained parameters for the model, wherein the second trained parameters have been generated by training the model using data from a second, different data cohort; determine a first evaluation value by inputting data from the first data cohort into a model having the first trained parameters; and determine a second evaluation value by inputting data from the first data cohort into a model having the second trained parameters.

Model training method and apparatus

A system including processing circuitry configured to train a model for predicting from input data at least one predicted output, wherein the processing circuitry is configured to: receive a plurality of training data sets; receive from a user a selection of a first characteristic including positive and negative samples which are relevant variations significant to prediction of the at least one predicted output; receive from the user a selection of a second characteristic including an irrelevant sample which is a spurious variation irrelevant to the prediction of the predicted output; perform positive supervision of the model using the first characteristic such that the training of the model is sensitive to the positive and negative samples of the first characteristic; and perform negative supervision of the model using the second characteristic such that the training of the model is insensitive to the irrelevant sample of the second characteristic.

Apparatus and method for training neural networks using small, heterogeneous cohorts of training data

An apparatus for determining similarity between medical data sets for a plurality of patients or other subjects comprises at least one data store and a processing resource. The at least one data store is configured to store a respective representation of each of a plurality of data sets, the representation of each data set being generated by applying a model for representing data sets with respect to a plurality of features. The processing resource is configured to use the model to obtain a representation of a further medical data set, and perform a similarity determining process to determine similarity between the representation of the further medical data set and at least some of said representations of said plurality of medical data sets.

Similarity determining apparatus and method

A medical image data processing apparatus comprises processing circuitry configured to: receive a plurality of sets of medical imaging data; and train a classifier for use in classification, wherein the training of the classifier comprises, for each of the plurality of sets of medical imaging data: selecting a first part and a second part of the respective set of medical imaging data, wherein the first part and the second part are representative of different regions of the same subject; and training the classifier for use in classification based on the first part of the set of medical imaging data and the second part of the set of medical imaging data.

Classification method and apparatus

A medical image processing apparatus comprises a memory configured to store medical image data representative of a tissue structure and a processing circuitry configured to operationally connect to the memory, extract regions from the medical image data by performing threshold processing of the medical image data using each of a plurality of threshold values, select regions meeting at least one predetermined condition from among the extracted regions, and determine a region representative of the tissue structure in the medical image data based on the selected regions.

Medical image processing apparatus and method

A medical imaging data processing apparatus comprises a receiving unit configured to receive first medical imaging data that represents a region of a subject at a first time, and second medical imaging data that represents the region of the subject at a second, later time, a registration unit configured to perform a registration procedure to obtain registration data representative of a registration between the first medical imaging data and the second medical imaging data, an evaluation unit configured to process the registration data to determine, for each of a plurality of positions in the region, a value of a registration measure, the registration measure representing at least one feature of the registration as a function of position, wherein the evaluation unit is configured to, for at least some of the plurality of positions in the region, determine whether an abnormality is present by comparing the value of the registration measure to a statistical atlas.

Medical image data processing apparatus and method for determining the presence of an abnormality

An image data processing apparatus including a data receiver receiving image data to be segmented, and an atlas selection processor accessing a plurality of atlas data sets and selecting a subset of the atlas data sets for use in segmenting the image data, wherein the atlas selection processor is configured to select the subset of atlas data sets in dependence on the positions of one or more anatomical landmarks comprised in the plurality of atlas data sets.

Apparatus and method for multi-atlas based segmentation of medical image data

Certain embodiments provide a computer apparatus operable to carry out a data processing method to position a set of anatomical landmarks in a three-dimensional image data set of a part or all of a patient, comprising: providing a trained supervised machine learning algorithm which has been trained to place each of the set of anatomical landmarks; applying the supervised machine learning algorithm to place the set of anatomical landmarks relative to the data set; providing a trained point distribution model, including a mean shape and a covariance matrix, wherein the mean shape represents locations of the set of landmarks in a variety of patients; and applying the point distribution model to the set of landmarks with the locations output from the supervised machine learning algorithm by: removing any landmarks whose locations have an uncertainty above a threshold with reference to the mean shape and covariance matrix; followed by: an optimization of the locations of the remaining landmarks by maximizing joint likelihood that a new shape, derived from linear combinations of eigenvectors of the covariance matrix, is plausible.

Positioning anatomical landmarks in volume data sets

A medical image processing apparatus comprises an image data processing unit configured to determine a result by processing image data using an algorithm, a confidence interval (CI) evaluation unit configured to determine a confidence interval for the result from at least two initial results, and an output unit configured to provide the confidence interval to a user.

Estimation of confidence limits for measurements derived from image data

An apparatus for locating a landmark in a set of image data comprises a landmark location unit that is configured, for each of a plurality of image data items, to obtain from a first two-class classifier a first classification of the image data item as foreground or background, to obtain from a second two-class classifier a second classification of the image data item as foreground or background, and to combine the first classification and the second classification to obtain a combined classification, and wherein the landmark location unit is further configured to use the combined classifications for the plurality of image data items to determine a location for the landmark.

Method of, and apparatus for, landmark location

A medical image processing apparatus, comprises a landmark identification unit configured to process a medical image data set to identify a landmark corresponding to a part of a branching structure according to a stored anatomical representation, and a spatial analysis unit configured to process the image data set to determine a spatial configuration of at least part of the branching structure.

Vascular tree from anatomical landmarks and a clinical ontology

Certain embodiments provide a computer system operable to determine a registration mapping between a first medical image and a second medical image, the computer system comprising: a storage device for storing data representing the first medical image and the second medical image; and a processor unit operable to execute machine readable instructions to: (a) identify a plurality of elements in the first medical image; (b) determine a spatial mapping from each element in the first medical image to a corresponding element in the second medical image to provide a plurality of spatial mappings subject to a consistency constraint; and (c) determine a registration mapping between the first medical image and the second medical image based on the plurality of spatial mappings from the respective elements of the first medical image to the corresponding elements of the second medical image.

Image registration

A method of processing image data including performing a first mapping, of a first image dataset to a first reference, performing a second mapping, of a second image dataset to a second reference, and using the first mapping and the second mapping to correlate at least one position in the first image dataset to at least one position in the second image dataset.

Image processing method and system

A method of locating anatomical features in a medical imaging dataset comprises obtaining a medical imaging measurement dataset that comprises image data for a subject body as a function of position; and performing a registration procedure that comprises:&#x2014;providing a mapping between positions in the measurement dataset and positions in a reference dataset, wherein the reference dataset comprises reference image data for a reference body as a function of position, the reference dataset comprises at least one anatomical landmark, and the or each anatomical landmark is indicative of the position of a respective anatomical feature of the reference body; matching image data in the measurement dataset with image data for corresponding positions in the reference dataset, wherein the corresponding positions are determined according to the mapping; determining a measure of the match between the image data of the measurement dataset and the image data of the reference dataset; varying the mapping to improve the match between the image data of the measurement dataset and the image data of the reference dataset, thereby to obtain a registration mapping; and using the registration mapping to map the positions of the anatomical landmarks to positions in the measurement dataset, thereby to assign positions to anatomical features in the measurement dataset.

Feature location method and system

According to one embodiment there is provided a computer-automated image processing method applied to a four-dimensional (4D) image data set of a patient's abdomen, e.g. by dynamic contrast enhanced computer-assisted tomography (DCE-CT). One of the three-dimensional (3D) scan images is taken to as the reference volume and the others as target volumes. Before registration between the 3D scan images, the image data set is partitioned into an abdominal cavity domain, containing the organs inside the abdominal wall, and an abdominal wall domain including the abdominal wall and externally adjacent skeletal features, such as the spine and ribs. Registration is then carried out separately on the two domains to obtain two warp fields which are then merged into a 4D image data set of the whole volume for further use, which may be to carry out perfusion measurements, to display and to store the registered 4D image data set.

Growing community of inventors

Edinburgh, United Kingdom

Ian Poole