Room: AAPM ePoster Library
Purpose: To compare a standard manual contouring workflow with two auto-contouring workflows (atlas and deep learning) for contouring the bladder and rectum in prostate cancer patients.
Methods: Three contouring workflows were defined based on the initial contour generation method including manual (MAN), atlas-based auto-contour (ATLAS) and deep learning auto-contour (DEEP). For each workflow, initial contour generation was retrospectively performed on fifteen prostate cancer patients. Then, three radiation oncologists (RO) edited each contour while blinded to the manner by which the initial contour was generated. Workflows were compared by time (both in initial contour generation and in RO editing) and geometric similarity of initial contours using Dice similarity coefficient (DSC) and mean surface separation. The extent of RO editing was also quantified by comparing initial and RO-edited contours using DSC and the fraction of contour points accepted (i.e. not edited by ROs).
Results: For initial contour generation, DEEP and ATLAS saved 9.5 min and 9.7 min respectively compared to MAN. Initial DEEP contours were more geometrically similar to initial MAN contours. Mean durations of the RO editing steps for MAN, DEEP and ATLAS contours were 4.1 min, 4.7 min and 10.2 min, respectively. The geometric extent of RO edits was consistently larger for ATLAS contours compared to MAN and DEEP.
Conclusion: Auto-contouring software affords time savings for initial contour generation; however, it is important to also quantify workload changes at the RO editing step. Using deep learning auto-contouring for bladder and rectum contour generation reduced contouring time without negatively affecting contour geometry or RO editing times. This work contributes to growing evidence that deep learning methods are a clinically viable solution for OAR contouring in radiotherapy.
Funding Support, Disclosures, and Conflict of Interest: Supported by Royal Victoria Hospital (RVH) Foundation.