System and method for image reconstruction

A non-transitory computer readable medium storing instructions, the instructions, when executed by a computer, causing the computer to implement a method comprising:

• receiving raw data relating to a subject;
• during each of a plurality of iterations,
• calculating first voxel values within a first voxel, second voxel values within a second voxel and third voxel values within a third voxel at each successive iteration based on an image obtained from a prior iteration, wherein the first voxel is adjacent to and shares a first common boundary with the second voxel along a first direction, the first voxel is adjacent to and shares a second common boundary with the third voxel along a second direction, the first voxel values include at least two different first values, the second voxel values include at least two different second values, and the third voxel values include at least two different third values, and wherein a first voxel value at the first common boundary is equal to a second voxel value at the first common boundary, and a first voxel value at the second common boundary is different from a third voxel value at the second common boundary;
• transforming the image to a projection domain to generate an estimated projection based on the first voxel values and the second voxel values;
• generating a projection error based on the estimated projection and the raw data;
• and
• generating an image based on the projection error to be used in a next iteration;
• and
• obtaining, based on the plurality of iterations, a reconstructed image.

An image processing system comprising:

• a data acquisition unit, implemented on at least one processor, configured to receive raw data relating to a subject;
• and
• an image reconstruction unit, implemented on the at least one processor, configured to perform operations comprising:
• performing a plurality of iterations;
• and during each of the plurality of iterations,
• calculating first voxel values within a first voxel, second voxel values within a second voxel and third voxel values within a third voxel at each successive iteration based on an image obtained from a prior iteration, wherein the first voxel is adjacent to and shares a first common boundary with the second voxel along a first direction, the first voxel is adjacent to and shares a second common boundary with the third voxel along a second direction, the first voxel values include at least two different first values, the second voxel values include at least two different second values, and the third voxel values include at least two different third values, and wherein a first voxel value at the first common boundary is equal to a second voxel value at the first common boundary, and a first voxel value at the second common boundary is different from a third voxel value at the second common boundary;
• transforming the image to a projection domain to generate an estimated projection based on the first voxel values and the second voxel values;
• generating a projection error based on the estimated projection and the raw data;
• and
• generating an image based on the projection error to be used in a next iteration;
• and
• obtaining, based on the plurality of iterations, a reconstructed image.