Fig 1.
Generation of DVFs utilizing VoxelMorph method.
(a) Generation of 10 modified CT images from 2 CT images of a patient to be employed as CTfixed images. (b) 10 DVFs extracted from each CTfixed via VoxelMorph method. Utilizing the 10 modified CT images and CT2nd image as a fixed image, 110 DVFs were extracted.
Fig 2.
CT label segmentation from deformed datasets.
(a) Deformed dataset generation through spatial transform of CT1st dataset with extracted DVF sets. (b) CT2nd label segmentation utilizing nn-Unet.
Fig 3.
Segmentation comparison using 3D volumes.
(A) Case 4 with the highest bladder (green colored organ) DSC of 0.969. (B) Case 14 with the lowest bladder DSC of 0.881. (C) Case 11 with the highest rectum (brown colored organ) DSC of 0.904. (D) Case 10 with the lowest rectum DSC of 0.753. (E) Case 3 with the highest prostate (yellow colored organ) DSC of 0.933. (F) Case 13 with the lowest prostate DSC of 0.673.
Fig 4.
Comparison of segmentation by overlapping CT and segmentation.
Case 01 contains air in the rectum; Case 07 contains no air in the rectum.
Table 1.
Comparison of DSC values with different networks, dataset numbers, VoxelMorph.
Table 2.
DSC, HD, and MSD performance evaluation of individual models in each patient.
Table 3.
DSC, HD, and MSD performance evaluation of total model in each patient.
Table 4.
Comparison of DSC values with state-of-the-art.
Fig 5.
Augmented CT images and labels generated in Case 04 utilizing VoxelMorph.
To verify whether the generated image exhibits a similar structure to the actual image, an examination of the image generated using VoxelMorph is presented for patient case 04.