| 摘要 |
[Problem] To carry out more precise alignment between two 3-D images obtained by photographing the same target organ of a patient at different respiratory phases. [Solution] A first 3-D image and second 3-D image obtained by photographing the same target organ of a patient at different respiratory phases are acquired. A 3-D deformation model for the target organ, which represents the non-linear 3-D deformation of the inside of the target organ due to respiration and has been generated and stored in advance in a storage means, is read. Using the displacement, due to the change in the respiratory phase, of points in the 3-D deformation model which correspond to positions on the target organ represented by each of a plurality of sampled pixels within the target organ region of the first 3-D image, the positions of the pixels in the second 3-D image which represent the same positions in the target organ as the sampled pixels are estimated. Non-rigid alignment is carried out between the first 3-D image and second 3-D image using the estimated positions of the pixels in the second 3-D image. |
