Abstract: A method and apparatus for selecting a subset of atlases from a set of candidate atlases. The method comprises: for each candidate atlas n within the set of N candidate atlases, computing an auto-contouring performance measure in relation to each training atlas t within a subset of a set of T training atlases; for each pair combination of candidate atlas n and training atlas t for which an auto-contouring performance measure has been computed, deriving a score based on the computed auto-contouring performance measures; for each candidate atlas n within the set of N candidate atlases, aggregating the derived scores for each pair combination comprising that candidate atlas n; and selecting a subset of M candidate atlases from the set of N candidate atlases based on the aggregated scores therefor.

Abstract: A method and apparatus for delineating an object within a volumetric medical image. The method comprises obtaining auto-generated contour data for the object within the volumetric medical image, the auto-generated contour data defining a set of auto-generated contours forming a delineation structure for the object, selecting a subset of auto-generated contours for manual editing, and identifying the selected subset of auto-generated contours to a user. In some examples, the method further comprises presenting at least the selected subset of auto-generated contours to the user, receiving user feedback for at least one auto-generated contour, deriving a full set of contours forming a revised delineation structure for the object based at least partly on an interpolation of the auto-generated contour(s) for which user feedback was received, and storing contour data defining the derived full set of contours forming the revised delineation structure for the object within at least one data storage device.

Abstract: A method and apparatus for selecting a subset of atlases from a set of candidate atlases. The method comprises: for each candidate atlas n within the set of N candidate atlases, computing an auto-contouring performance measure in relation to each training atlas t within a subset of a set of T training atlases; for each pair combination of candidate atlas n and training atlas t for which an auto-contouring performance measure has been computed, deriving a score based on the computed auto-contouring performance measures; for each candidate atlas n within the set of N candidate atlases, aggregating the derived scores for each pair combination comprising that candidate atlas n; and selecting a subset of M candidate atlases from the set of N candidate atlases based on the aggregated scores therefor.

Abstract: A method and apparatus for delineating an object within a volumetric medical image. The method comprises obtaining auto-generated contour data for the object within the volumetric medical image, the auto-generated contour data defining a set of auto-generated contours forming a delineation structure for the object, selecting a subset of auto-generated contours for manual editing, and identifying the selected subset of auto-generated contours to a user. In some examples, the method further comprises presenting at least the selected subset of auto-generated contours to the user, receiving user feedback for at least one auto-generated contour, deriving a full set of contours forming a revised delineation structure for the object based at least partly on an interpolation of the auto-generated contour(s) for which user feedback was received, and storing contour data defining the derived full set of contours forming the revised delineation structure for the object within at least one data storage device.