PURPOSE: A method is provided for the automatic calculation of the Clinical Target Volume (CTV) by automatic three dimensional (3D) expansion of the Gross Tumor Volume (GTV), keeping a constant margin M in all directions and taking into account anatomic obstacles. METHODS AND MATERIALS: Our model uses a description of the GTV from contours (polygons) defined in a series of parallel slices obtained from Computed Tomography (CT) or Magnetic Resonance Imaging (MRI). Each slice is considered sequentially, including those slices located apart from the GTV at a distance smaller than M. In the current slice a two-dimensional (2D) expansion is performed by transforming each vertex of the polygon into a circle with a radius equal to M, and each segment into a rectangle with a height equal to 2M. A cartesian millimetric grid is then "projected" onto the slice and a specific value is assigned at each point depending if the point is internal to the 2D expansion. The influence in the current slice of any slice located at a distance delta z smaller than M is taken into account by applying a 2D expansion using a margin [formula: see text]. Additional contours representative of various "barriers" stopping the expansion process can also be defined. RESULTS: The method has been applied to cylindrical and spherical structures and has proven to be successful, provided that the slice thickness is small enough. For usual slice thicknesses and margins, it gives a slight overestimation of the additional volume (around 5%) due to the choice that the calculated target volume would not be less than the expected volume. It has been shown that for a spherical volume, a 2D expansion performed slice by slice leads to a volume up to 80% smaller than that obtained by 3D expansion. CONCLUSIONS: This tool, which mimics the tumor cell spreading process, has been integrated in our treatment-planning software and used clinically for conformal radiotherapy of brain and prostatic tumors. It has been found to be extremely useful, not only saving time but also allowing a precise determination of the CTV which would be impossible to do manually.
PURPOSE: A method is provided for the automatic calculation of the Clinical Target Volume (CTV) by automatic three dimensional (3D) expansion of the Gross Tumor Volume (GTV), keeping a constant margin M in all directions and taking into account anatomic obstacles. METHODS AND MATERIALS: Our model uses a description of the GTV from contours (polygons) defined in a series of parallel slices obtained from Computed Tomography (CT) or Magnetic Resonance Imaging (MRI). Each slice is considered sequentially, including those slices located apart from the GTV at a distance smaller than M. In the current slice a two-dimensional (2D) expansion is performed by transforming each vertex of the polygon into a circle with a radius equal to M, and each segment into a rectangle with a height equal to 2M. A cartesian millimetric grid is then "projected" onto the slice and a specific value is assigned at each point depending if the point is internal to the 2D expansion. The influence in the current slice of any slice located at a distance delta z smaller than M is taken into account by applying a 2D expansion using a margin [formula: see text]. Additional contours representative of various "barriers" stopping the expansion process can also be defined. RESULTS: The method has been applied to cylindrical and spherical structures and has proven to be successful, provided that the slice thickness is small enough. For usual slice thicknesses and margins, it gives a slight overestimation of the additional volume (around 5%) due to the choice that the calculated target volume would not be less than the expected volume. It has been shown that for a spherical volume, a 2D expansion performed slice by slice leads to a volume up to 80% smaller than that obtained by 3D expansion. CONCLUSIONS: This tool, which mimics the tumor cell spreading process, has been integrated in our treatment-planning software and used clinically for conformal radiotherapy of brain and prostatic tumors. It has been found to be extremely useful, not only saving time but also allowing a precise determination of the CTV which would be impossible to do manually.