A Parallel Nonrigid Registration Algorithm Based on B-Spline for Medical Images

Algorithm 3

The pseudocode listing of the kernel function CalSimilarityGradientKernelFunc().

 Get the block index and the local thread index .

 for to do

    ;

    Get the normalized coordinates for the voxel within the region.

    ;

    for from to do

     ;

     ;

     Reduce and store in

    _syncthreads();

    if then

       ;

    end if

    _syncthreads();

    ;

    end for

  end for

  ;

  ;

Use tree-style reduction to reduce the gradient values in the shared memory.

Store the reduction results in the global memory of GPU.