Development and Implementation of an Image Registration and Fusion Workshop for Radiation Oncology Trainees

BackgroundModern radiation oncology relies on image registration and fusion to accomplish daily tasks. These tasks include 1) contouring target volumes, 2) contouring organs at risk for treatment planning, and 3) daily verification of image guided radiation therapy. Despite the reliance of these techniques to ensure high quality treatment planning and delivery, literature suggests there is currently no standard education for trainees to understand, create, or verify image registrations. Our purpose was to create and implement an image registration and fusion workshop to fill gaps in trainee confidence, skills, and knowledge.MethodsThe voluntary image registration workshop consisted of a hands-on portion where participants independently performed image registrations and a didactic portion containing lectures presented by a medical physicist and chief resident. All trainees at our institution were invited to participate. Participants filled out pre- and post-workshop surveys which evaluated their confidence in creation, evaluation, and understanding of image registrations. Pre- and post-workshop registration assignments - organized by five major anatomical disease sites - were created and quantitatively compared (i.e., magnitude of rigid registration translations differences) to previously clinically approved fusions. Deviations compared to clinically accepted registrations, group similarity between registrations, and pre- and post-survey differences were evaluated using a Wilcoxon rank-sum test. All participant assignments were also evaluated independently by a blinded attending physician and physicist to determine if the fusion would be clinically approved.ResultsSeven trainees agreed to participate (five radiation oncology residents, two medical physics residents). Pre- and post-workshop surveys were completed by all participants. Significant increases (p < 0.05) in participant confidence were observed for every survey question. In the head and neck anatomical subgroup, there was a significant decrease in the translational differences between participant and clinically approved registrations pre- and post-workshop (median differences: 10.15 vs 5.64 mm, pre and post; p < 0.01). No significant differences were noted in other disease sites. The range of differences between participant registrations compared to clinically approved fusions for pre- and post-workshop assignments were 64.49 and 9.32 mm, respectively. Physician and physicist evaluation of fusions revealed no significant differences in clinical acceptability between pre-and post-workshop assignments.DiscussionAn image registration and fusion curriculum improved participant confidence in evaluating, creating, and understanding registrations for five major disease sites and decreased the magnitude of outliers in participant fusions. We plan to continue curriculum development and interprofessional educational opportunities to make trainees more confident and competent with image registrations and fusions.