Wisconsin Alumni Research Foundation

Hyungseok Jang

Alan McMillan

Richard Kijowski

Tyler Bradshaw

Fang Liu

The present disclosure includes systems and methods for creating positron emission tomography (PET) images. The method includes receiving at least one PET image of a subject created from PET data acquired from the subject, creating an attenuation correction map using the at least one PET image, and reconstructing PET data using the attenuation correction map and the at least one PET image to generate an attenuation corrected PET image.

Deep learning based data-driven approach for attenuation correction of pet data

A system and method for determining an actual gradient field generated by a magnetic resonance imaging (MRI) system when controlled to produce a prescribed gradient field is provided. The techniques include using the prescribed gradient field, controlling the MRI system to perform a phase encoding including a gradient that is scaled along each direction desired to be measured over a selected number of encoding times and acquiring one-dimensional (1D) data using a prescribed k-space trajectory during the phase encoding. The 1D data is used to determine scaling factors between encoding times that correlate to actual k-space trajectories achieved when controlling the gradient coils to perform the phase encoding based on the desired gradient field and a report is generated that provides a measure of the actual gradient field generated when controlling the MRI system to produce the prescribed gradient field.

Growing community of inventors

Madison, WI, United States of America

Hyungseok Jang