1. For each training MR image patch , extract the multiscale and multilevel 3-D RILBP feature . |
2. Use fuzzy C-means to segment each corresponding CT into three labels that represent the CT values in the range of bone, soft-tissue and air. |
3. For the coupled training MRI features and corresponding CT labels, apply LASSO logistic regression Eq. (1) to select the informative features as an anatomic signature. |
4. Use Fisher’s score47 to measure the discriminant power of the anatomic signature to separate bone, air and soft-tissue, and normalize to such that . |
5. For each new arrival MRI image patch , select its similar patches from the training patches by computing the weight from Eq. (2), and sorting the K-nearest patches by Eq. (3). |
6. For the anatomic features of and corresponding CT voxels , initialize the coupled dictionaries: and . |
7. Use joint dictionary learning [Eq. (4)] to train the coupled dictionary . |
8. For each new arrival MRI patch, use Eq. (5) to obtain the sparse representation of learned dictionary , and then use Eq. (6) to reconstruct the pseudo CT intensity. |