Method and apparatus for node-based motion estimation in cardiac CT imaging

1 . A method for motion estimation in a CT system, comprising:

dividing projection data, obtained by scanning a heart using the CT system, into a plurality of partial-angle-reconstruction (PAR) bins;

reconstructing a plurality of PAR volumes from the PAR-binned projection data;

obtaining, based on the plurality of reconstructed PAR volumes, a number of short-scan volumes;

determining, based on the obtained number of short-scan volumes, a plurality of nodes throughout the heart;

estimating, for each of the determined plurality of nodes, a plurality of model parameters of a motion model; and

generating, based on the plurality of model parameters estimated for each of the plurality of nodes, parameters of a global motion model at each voxel of a volume of the heart.

2 . The method of claim 1 , further comprising:

reconstructing, based on the generated motion parameters of the global motion model at each voxel of the volume of the heart, a motion-compensated short-scan volume.

3 . The method of claim 1 , wherein the determining step further comprises:

determining, based on the obtained number of short-scan volumes and additional information, the plurality of nodes throughout the heart, wherein

the additional information includes a segmented coronary artery tree.

4 . The method of claim 3 , wherein the determining step further comprises:

determining, based on the obtained number of short-scan volumes and the additional information, a plurality of artery nodes, a starting node, and a plurality of other region nodes throughout the heart, wherein

the plurality of artery nodes are distributed along the segmented coronary artery tree,

the starting node is determined to be located in an anatomical area known to have motion less than a threshold, and

the plurality of other region nodes are distributed based on a comparison of the number of short-scan volumes.

5 . The method of claim 4 , wherein the determining step further comprises:

creating a node sequence including the plurality of artery nodes, the starting node, and the plurality of other region nodes, by placing the starting node at a first slot of the node sequence, and sorting the plurality of other region nodes such that consecutive nodes in the node sequence are close to each other spatially.

6 . The method of claim 3 , wherein the additional information further includes a heart mask such that the plurality of node are determined to be only inside the heart.

7 . The method of claim 1 , wherein the number of short-scan volumes include a first short-scan volume, a second short-scan volume, and a reference short-scan volume;

the reference short-scan volume is centered at a reference phase where motion is assumed to be minimal;

both the first short-scan volume and the second short-scan volume are centered away from the reference phase; and

the estimating step further comprises, for each of the plurality of nodes, determining the plurality of model parameters of the motion model so as to maximize a cost functional that includes a similarity between the first short-scan volume corrected with the motion model and the second short-scan volume corrected with the motion model.

8 . The method of claim 1 , wherein the cost functional further includes a sharpness term that is adjusted based on whether a depressed Hounsfield Unit value is identified as contributing to a sharpness measurement of the reference short-scan volume.

9 . The method of claim 1 , wherein the estimating step further comprises estimating, for each of the plurality of nodes, the plurality of model parameters of the motion model, which are determined by analyzing a local volume around the node, and

a size of the local volume is determined based on an orientation of a blood vessel where the node is located.

10 . The method of claim 1 , wherein the estimating step further comprises, for each of the plurality of nodes:

estimating the plurality of model parameters of the motion model, which are determined by analyzing a local volume around the node, and

upon detection, in a vicinity of the node, of a feature that has a larger sharpness measurement than the node, moving the node to a position of the detected feature, and updating the local volume to cover both the original local volume around the initial node and a local volume around the detected feature.

11 . The method of claim 1 , wherein the dividing step further comprises:

dividing the projection data into the plurality of PAR bins, wherein every pair of adjacent PAR bins has an overlap represented by a partition-of-unity weight.

12 . An apparatus for motion estimation in a CT system, the apparatus comprising: processing circuitry configured to

divide projection data, obtained by scanning a heart using the CT system, into a plurality of partial-angle-reconstruction (PAR) bins;

reconstruct a plurality of PAR volumes from the PAR-binned projection data;

obtain, based on the plurality of reconstructed PAR volumes, a number of short-scan volumes;

determine, based on the obtained number of short-scan volumes, a plurality of nodes throughout the heart;

estimate, for each of the determined plurality of nodes, a plurality of model parameters of a motion model; and

generate, based on the plurality of model parameters estimated for each of the plurality of nodes, parameters of a global motion model at each voxel of a volume of the heart.

13 . The apparatus of claim 12 , where the processing circuitry is further configured to:

reconstruct, based on the generated motion parameters of the global motion model at each voxel of the volume of the heart, a motion-compensated short-scan volume.

14 . A non-transitory computer-readable medium storing a program that, when executed by processing circuitry, causes the processing circuitry to execute a method for motion estimation in a CT system, the method comprising:

dividing projection data, obtained by scanning a heart using the CT system, into a plurality of partial-angle-reconstruction (PAR) bins;

reconstructing a plurality of PAR volumes from the PAR-binned projection data;

obtaining, based on the plurality of reconstructed PAR volumes, a number of short-scan volumes;

determining, based on the obtained number of short-scan volumes, a plurality of nodes throughout the heart;

estimating, for each of the determined plurality of nodes, a plurality of model parameters of a motion model; and

generating, based on the plurality of model parameters estimated for each of the plurality of nodes, parameters of a global motion model at each voxel of a volume of the heart.

15 . The non-transitory computer-readable medium of claim 14 , wherein the method further comprises:

reconstructing, based on the generated motion parameters of the global motion model at each voxel of the volume of the heart, a motion-compensated short-scan volume.

16 . The method of claim 1 , wherein the determining step further comprises determining, by analyzing features in the obtained number of short-scan volumes, the plurality of nodes throughout the heart; and

the estimating step further comprises estimating, for each of the determined plurality of nodes, by analyzing the features in the obtained number of short-scan volumes in a neighborhood of each node, a plurality of model parameters of a motion model of that node.