Skip to content

dir_validation_data.md

Latest commit

 

History

History
57 lines (29 loc) · 5.34 KB

dir_validation_data.md

File metadata and controls

57 lines (29 loc) · 5.34 KB
Raw

Validation Data for Deformable Image Registration

Motivation

We believe research on deformable image registration (DIR) of the lungs will greatly benefit from making image and validation data publicly available. Several 4D CT data sets, acquired at the Léon Bérard Cancer Center, can be downloaded here for use in your research. Along with the images, we provide large sets of manually identified landmarks, allowing you to validate your motion estimation.

We welcome all suggestions that can make this data more useful for you. David Sarrut, david.sarrut[at]creatis.insa-lyon.fr

Downloads

By downloading this data, you agreeing to follow these guidelines. When using any of the data provided for publications, please make reference to:

  • The institution from which they were obtained: the Léon Bérard Cancer Center & CREATIS lab, Lyon, France.

  • The article containing the description of this data:

    J. Vandemeulebroucke, S. Rit, J. Kybic, P. Clarysse, and D. Sarrut. Spatiotemporal motion estimation for respiratory-correlated imaging of the lungs. In Med Phys, 2011, 38(1), 166-178. pdf

The inhale-exhale image pairs were also used in

  • J. Vandemeulebroucke, O. Bernard, S. Rit, J. Kybic, P. Clarysse, and D. Sarrut. Automated segmentation of a motion mask to preserve sliding motion in deformable registration of thoracic CT. In Med Phys, 2012, 39(2), 1006-1015. pdf

The landmarks were identified using a software tool proposed by Keelin Murphy

  • K. Murphy, B. van Ginneken, S. Klein, M. Staring, B. de Hoop, M. Viergever & J. Pluim. Semi-automatic construction of reference standards for evaluation of image registration. In Med Im Anal, 2011, 15, 71-84.
Patient Dicom Files Reference Points Reference Phase MHD Files Translation (Add this value to coordinates y for points to display them on the dicom, in mm)
1 Images Points 60 Images 85.9375
2 Images Points 50 Images 54.6875
3 Images Points 50 Images 54.4921875
4 Images Points 50 Images 55.46875
5 Images Points 50 Images 55.078125
6 Images Points 50 Images 55.078125

Data Description

Images were acquired on a 16 Slice Brilliance CT Big Bore Oncology™ configuration (Philips). Breathing correlated information was obtained using the associated Pneumo Chest bellows™ (Lafayette Instruments). Each data set consists of ten 3D CT frames, respresenting a breathing cycle.

The landmarks were identified using a software tool proposed by Keelin Murphy (for a full description of the identification procedure followed, see [Vandemeulebroucke2012.pdf]). For the first three patients, approximately a hundred landmarks were identified in all frames. For the remaining three patients, a hundred landmarks were provided in the inhale and exhale frame.

Landmarks are defined in the MHD coordinate system. The coordinates in the Dicom Images are given in the CT scanner coordinate system while MHD images use the coordinate system of our Treatment Planing System that puts the origin at the center of the image in the Y direction compensating for the hight of the CT couch. To display landmarks in the Dicom Images, a Y translation must be applied, either to the Dicom Images or to the landmarks themselves, so that the origin is centered in the middle of the CT image in the Y direction.

Other Useful Links