Mesh based ROI correction interface for organ delineation in radiation oncology planning

Organ delineation from volumetric images is an integral part of radiation oncology planning. Model based segmentation techniques have been successfully applied for automatic segmentation of region of interest (ROI) in volumetric medical images. The ROI is modelled as a 3D triangular surface mesh which automatically adapts itself to fit to the correct boundary of the ROI in a given image. Automatic segmentation could be inaccurate due to attraction to false boundaries, which can happen because of poor model initialization or weak image feature response at correct locations. To be acceptable in clinical practice, it is crucial for a segmentation approach to integrate corrective input from the user. A novel mesh editing technique is discussed in this work. The editing interface provided is intuitive and simple; the user draws a free hand curve to mark the correct boundary of the ROI. The proposed mesh editing algorithm uses distance and curvature metrics to identify and deform a subset of mesh vertices with respect to the user curve. The neighbouring vertices that are within the user defined depth get deformed based on shape constraint. The shape constraint is formulated to maintain the distribution of the vertices in the region of edited mesh in correspondence with the initial surface model. Mesh editing results accurately (0.38769 ± 0.19779 mm) match with the user desired corrections on the viewed image plane. The shape of neighbouring region is also preserved resulting in smooth and regular contours on adjacent image planes.