Shannon Costello Abstract
Shannon Costello, Medical Physics
Dr. Diana Shvydka and Dr. David Pearson
Atlas Based Automatic Segmentation in a Radiation Therapy Setting
Radiation Oncology treatment planning is an elaborate process that includes multiple steps and reviews before a plan is approved for delivery. One of the most time-consuming tasks during the process is contouring or outlining, the patient’s organs at risk on computed tomography (CT) scan in order to prevent unnecessary radiation exposure to healthy tissue. In addition to the time concern, there is also a significant amount of intra- and inter-observer inconsistencies during the contouring procedure: the different clinicians may interpret the boundaries of each organ differently. A solution to these two hurdles incorporates the implementation of an atlas-based automatic contouring software. The accuracy of such an approach has to be evaluated to determine if the software is able to produce contours similar to those manually drawn, which are used as the “gold standard” contours. For this purpose, atlases were created in various portions of the body using MIM software package and applied to previously treated and already drawn contours on patient CTs. The individual organs of both atlas-based contours and manually drawn contours were compared volumetrically with the Dice Similarity Coefficient (DSC) ranging from 0, meaning no volumetric overlap, to 1, meaning complete volumetric overlap. The average DSC value for all software-generated contours was found to be 0.674813 for all regions considered, a modest comparison between automatically generated segmentations and manually generated segmentations.