Accelerated iterative reconstruction
A framework for an iterative reconstruction algorithm is described which combines two or more of an ordered subset method, a preconditioner method, and a nested loop method. In one type of implementation a nested loop (NL) structure is employed where the inner loop sub-problems are solved using ordered subset (OS) methods. The inner loop may be solved using OS and a preconditioner method. In other implementations, the inner loop problems are created by augmented Lagrangian methods and then solved using OS method.
1. A processor-implemented image reconstruction method, the method comprising:
on an imaging system or a remote client in communication with the imaging system, for a given set of scan data acquired by the imaging system determining an objective function of an inner loop nested within an outer loop;
iteratively processing the inner loop using ordered subsets until the inner loop is determined to be complete, wherein iteratively processing the inner loop comprises:
applying a preconditioner; and
computing an image update;
upon completion of a respective inner loop:
computing an update direction for the respective outer loop, wherein the update direction is determined by combining a plurality of inner loop update directions from the inner loops encompassed by the respective outer loop;
computing a step size for the respective outer loop; and
updating an image;
determining if the image is converged:
if the image is not converged, proceeding to a next outer loop, determining a next objective function of a next inner loop, and iteratively processing the next inner loop; and
if the image is converged, ending the image reconstruction, generating a final image, and displaying a final image on a display of the imaging system or the remote client.
2. The image reconstruction method of claim 1 , wherein the step size for each respective outer loop is computed to minimize the objective function associated with the respective outer loop.
3. The image reconstruction method of claim 1 , wherein the preconditioner is a combination of two or more other preconditioners.
4. The image reconstruction method of claim 1 , wherein gain associated with each component of the preconditioner are adapted each iteration.
5. An image processing system for use in iterative reconstruction, comprising:
a processing component of an imaging system or a remote client in communication with the imaging system, wherein the processing component is configured to access acquired image data from the imaging system; and
a memory configured to store one or more routines which, when executed by the processing component, cause acts to be performed comprising:
for a given set of acquired image data, determining an objective function of an inner loop nested within an outer loop;
iteratively processing the inner loop using ordered subsets until the inner loop is determined to be complete, wherein iteratively processing the inner loop comprises:
applying a preconditioner; and
computing an image update;
upon completion of a respective inner loop:
computing an update direction for the respective outer loop,
wherein the update direction is determined by combining a plurality of inner loop update directions from the inner loops encompassed by the respective outer loop;
computing a step size for the respective outer loop; and
updating an image;
determining if the image is converged:
if the image is not converged, proceeding to a next outer loop, determining a next objective function of a next inner loop, and iteratively processing the next inner loop; and
if the image is converged, ending the image reconstruction, generating a final image, and displaying the final image on a display of the imaging system or the remote client.
6. The image processing system of claim 5 , wherein the step size for each respective outer loop is computed to minimize the objective function associated with the respective outer loop.
7. The image processing system of claim 5 , wherein the preconditioner is a combination of two or more other preconditioners.
8. The image processing system of claim 5 , wherein gains associated with each preconditioner component are adapted each iteration.
9. One or more non-transitory computer-readable media encoding routines which, when executed by a processor of an imaging system or a remote client in communication with the imaging system, cause acts to be performed comprising:
for a given set of scan data acquired by the imaging system, determining an objective function of an inner loop nested within an outer loop;
iteratively processing the inner loop using ordered subsets until the inner loop is determined to be complete, wherein iteratively processing the inner loop comprises:
applying a preconditioner; and
computing an image update;
upon completion of a respective inner loop:
computing an update direction for the respective outer loop wherein the update direction is determined by combining a plurality of inner loop update directions from the inner loops encompassed by the respective outer loop;
computing a step size for the respective outer loop; and
updating an image;
determining if the image is converged;
if the image is not converged, proceeding to a next outer loop, determining a next objective function of a next inner loop, and iteratively processing the next inner loop; and
if the image is converged, ending the image reconstruction, generating a final image, and displaying the final image on the display.
10. The one or more non-transitory computer-readable media of claim 9 , wherein the step size for each respective outer loop is computed to minimize the objective function associated with the respective outer loop.
11. The one or more non-transitory computer-readable media of claim 9 , wherein the preconditioner is a combination of two or more other preconditioners.
12. The one or more non-transitory computer-readable media of claim 9 , wherein gains associated with each preconditioner component are adapted each iteration.