Ge Medical Systems Global Technology Company, LLC

General Electric Company

Zhanfeng Xing

Jiang Hsieh

Bruno Kristiaan Bernard De Man

Bernhard Erich Hermann Claus

Yannan Jin

A method relates to the use of deep learning techniques, which may be implemented using trained neural networks (), to estimate various types of missing projection or other unreconstructed data. Similarly, the method may also be employed to replace or correct corrupted or erroneous projection data as opposed to estimating missing projection data.

Deep learning based estimation of data for use in tomographic reconstruction

A gantry rotation control device for a computed tomography scanning apparatus. The gantry rotation control device includes a radiation dose determination unit for determining radiation doses of the X-rays that will be emitted to each site of the target object to be scanned, a minimum velocity determination unit for determining a minimum rotation velocity of the gantry according to a maximum radiation dose in the determined radiation doses, a maximum velocity determination unit for determining a maximum rotation velocity of the gantry according to the determined minimum rotation velocity, a rotation velocity determination unit for determining a rotation velocity of the gantry at any time during scanning of the target object according to the determined minimum rotation velocity and maximum rotation velocity, and a gantry rotation control unit for controlling the gantry to scan the target object while rotating according to the determined rotation velocity when the target object is to be scanned.

Computed tomography scanning apparatus, gantry rotation control device and method thereof

An object of the present invention is to provide an image processing method for reducing noises without causing a side effect and an X-ray CT system that performs the image processing. An input image is wavelet-transformed using a gradient form. Resultant wavelet vectors are classified into categories of strongly edged objects, weakly edged objects, radial lines, and noises. The wavelet vectors classified into the category of strongly edged object are left intact. The wavelet vectors classified into the category of weakly edged objects have the directions thereof filtered. The wavelet vectors classified into the categories of radial lines and noises are smoothed. The resultant wavelet vectors are inverse-wavelet-transformed in order to produce an output image.

Growing community of inventors

Beijing, China

Zhanfeng Xing