Fig 1.
An example of intra-operative image, note the rotation of the patient required for safe needle insertion from the side.
Fig 2.
Registration scheme of the diagnostic contrast enhanced CT image D and the intra-operative CT images: Stage 1 is registration between the image D and the full operative image F; and stage 2 is the registration between the image D and the operative image N.
Fig 3.
Example of incorrect deformation at the edges of the liver in case the liver boundary in the intra-operative image is unclear.
The arrow points to the region with incorrect deformation.
Fig 4.
The coefficient mask (in red) is used to penalize the rigid deformation of the liver inside the liver mask (in yellow).
Fig 5.
The manual liver segmentation and the landmarks inside the liver.
Table 1.
Registration evaluation 16 cases in the first stage.
Fig 6.
Two examples of successful registrations between diagnostic image and intra-operative image with the tumor (in red): The top row shows the diagnostic images; the second row shows the intra-operative images; the third row shows the fused images. The left column is an example of a contrast enhanced intra-operative image and the right column is an example of a non-contrast enhanced intra-operative image with 30 degree-rotation. The registration method only computes the metric within the liver, thus the images do not match outside the liver.
Fig 7.
The effect of the rigidity term (from Eq 1) a) Registration without rigidity constrain, yielding large incorrect deformation (red arrow). b) Registration with rigidity constrain in the refined stage.
Fig 8.
Example of an unsuccessful registration.
Fig 9.
Example of tumor visualization in the intra-operative image after registration to the diagnostic image with the tumor (in red): a) the initial intra-operative image; b) the transformed diagnostic image to the initial intra-operative image; c) the next fame of intra-operative image; d) the transformed diagnostic image to the current intra-operative image.
Table 2.
Registration evaluation of 16 cases in the second stage.