Below is a description of the characteristics of each of the currently available reference data sets. Each of cases c1 - c5 were first cropped to include the entire ribcage, then sub-sampled in-plane to 256 x 256 voxels. The smaller image volume dimensions are included to ensure that the largest possible range of DIR algorithms is able to be implemented on at least some subset of reference data sets (for example, in some instances computing resources may not be available to determine a DIR solution for a full resolution 512 x 512 x N volume pair). For c1 - c5, the original in-plane voxel dimensions were 0.97 mm x 0.97 mm. Cases 6 - 10 were not subsampled or processed in any way.

The entries in the "# Repeated Measurements" column are formatted as (N_m / N_obs), where N_m is the number of repeated registration measurements for the given case, performed by each of N_obs independent observers. For each case, the "Observer Error" column shows the combined mean repeated registration error for the set of N_obs data sets. The "Landmarked Phases" entries indicate the component phase images for which manually registered landmarks have been identified. T00 represents the maximum inhale phase of the 4DCT, while T50 represents the maximum exhale phase. We will update the available reference data as more cases are added, and as the current sets of landmark correspondences are propagated onto additional component phases for each case. Furthermore, 'BH-CT' indicates image pairs obtained from breath-hold CT.

For the 4DCT data sets, cine loops through representative axial, coronal, and sagittal views are also shown below for each case. In addition, surface renderings of the inhale lung volumes are shown, overlain with displacement vectors (T00 --> T50) for each of the validation landmarks. All surface renderings were generated in the MATLAB (Mathworks, Sunnyvale, CA) environment.

Representative cine loops (4DCT only) and surface renderings are shown below for all available reference data sets listed in Table 1 (click link to view).

Case 1:

Axial Cine Loop
Coronal Cine Loop
Sagittal Cine Loop

Back to top

Case 2:

Axial Cine Loop
Coronal Cine Loop
Sagittal Cine Loop

Back to top

Case 3:

Axial Cine Loop
Coronal Cine Loop
Sagittal Cine Loop

Back to top

Case 4:

Axial Cine Loop
Coronal Cine Loop
Sagittal Cine Loop

Back to top

Case 5:

Axial Cine Loop
Coronal Cine Loop
Sagittal Cine Loop

Back to top

Case 6:

Axial Cine Loop
Coronal Cine Loop
Sagittal Cine Loop

Back to top

Case 7:

Axial Cine Loop
Coronal Cine Loop
Sagittal Cine Loop

Back to top

Case 8:

Axial Cine Loop
Coronal Cine Loop
Sagittal Cine Loop

Back to top

Case 9:

Axial Cine Loop
Coronal Cine Loop
Sagittal Cine Loop

Back to top

Case 10:

Axial Cine Loop
Coronal Cine Loop
Sagittal Cine Loop

Back to top