Movement correction in MRI using a camera,申请号US200912937852-传众专利搜索

Movement correction in MRI using a camera

申请公布号：US9545217(B2)

申请号：US200912937852

申请日期：2009.04.17

申请公布日期：2017.01.17

申请人：

The United States of America, as Represented by the Secretary, Department of Health and Human Services

发明人：Duyn Jeff;Qin Lei

分类号：G01R33/567;A61B5/055;A61B5/05;G01R33/565;G06T7/00;G06T3/00;A61B5/00;G01R33/28;G01R33/48;G01R33/54

主分类号：G01R33/567

代理机构：

Polsinelli PC

代理人：Polsinelli PC ;Bai Ari M.

地址：Washington DC US

摘要：Provided are methods and systems for movement correction in an MRI environment. In one aspect, provided are systems and methods for movement correction, comprising receiving a first plurality of images from a first scan of a subject with a first camera, receiving magnetic resonance imaging (MRI) images obtained concurrently with the first scan, correlating the first plurality of images obtained from the first scan with the MRI images, resulting in motion correction data, and providing the motion correction data to an MRI system, wherein the MRI system adjusts scanning according to the motion correction data.

主权项：1. A method for motion correction in a magnetic resonance imaging (MRI) system, comprising: generating a training scan, comprising the steps of: receiving a plurality of training camera images from a training scan of a subject or a portion thereof obtained by a camera;receiving a plurality of training MRI images of the subject or portion thereof obtained by an MRI device, the plurality of training MRI images being obtained concurrently with the plurality of training camera images, wherein each one of the plurality of training camera images corresponds to a respective one of the plurality of training MRI images as a function of time;using a processor to perform the steps of: extracting one or more training image features from each of the plurality of training camera images;using an image registration algorithm to derive motion parameters from the plurality of training MRI images, the motion parameters being parameters in a three-dimensional MRI coordinate system; andgenerating a translation table in which each entry in the translation table assigns to each of the plurality of training camera images the motion parameters derived from each corresponding concurrent training MRI image of the plurality of training MRI images;generating a testing scan using the processor to perform the steps of: receiving a plurality of testing camera images of the subject or portion thereof obtained with the camera;extracting one or more testing image features of the subject or a portion thereof from the testing camera images;comparing the testing camera images with the training camera images to find the most similar images between the training camera images and the testing camera images, wherein the position of each testing image feature on the subject or portion thereof corresponds to a training image feature, andcalculating the distance between the one or more testing image features and the corresponding one or more training image features to find the minimum distance between the one or more testing image features and the one or more training image features, wherein the one or more testing image features and the one or more training image features are locations in a physical space, wherein the summation of distance between the one or more training and testing image features in the physical space indicates the similarity thereof; andreceiving a plurality of testing MRI images obtained by the MRI device, the plurality of testing MRI images being obtained concurrently with the plurality of testing camera images, wherein each testing camera image corresponds to a testing MRI image as a function of time; andapplying a motion correction to the plurality of testing MRI images by providing the data from translation table to the MRI system, wherein the MRI device adjusts MRI scanning parameters comprising gradient rotation and offset phases and frequencies in the MRI scanner sequence in real-time according to motion parameters in the translation table that are assigned to the training camera image that is most similar to the testing camera image to account for motion of the subject or portion thereof during the testing scan.