INTERACTIVE REGION AND VOLUME GROWING FOR SEGMENTING VOLUMES IN MR AND CT IMAGES

In radiation treatment planning it is necessary to mark out selected p arts of CT or MR images which are clinically relevant before designing the treatment plan. This process is usually done manually with a mous e, and for a typical data set of some 40 slices it can be a time-consu ming and tedious process. We report on a system for interactive region and volume growing that assists the clinical user in outlining releva nt volumes. The algorithm detects regions of interest grown from a see d selected interactively by a user. The regions consist of connected a reas or volumes having similar pixel intensity values. Threshold param eters for controlling the algorithm can be set interactively, or may b e determined automatically from a sample region. Region growth can be successively extended into adjacent slices so building up a three-dime nsional (3D) volume. Direct 3D volume growth from a seed voxel is also possible. The algorithm has been integrated into one of the pilot sys tems developed as part of COVIRA, project A2003 in the EC's AIM progra mme.