Nth Slice Eraser: An Automated Algorithm to Ease Editing Workflow of Organ Contours Generated by Artificial Intelligence

The time-consuming process of manual contouring of healthy tissue and organs in radiation therapy has prompted the development of computational systems to aid and automate this process, such as artificial intelligence (AI) segmentation and interpolation algorithms. These algorithms are useful in saving time, however, they are not always accurate. Fixing such inaccuracies by editing contours is a manual, time-consuming process as no 'undo' feature currently exists in the most commonly used treatment planning system (TPS). In this short technical report, we present our script for automated de-interpolation to enable faster editing of contours to streamline the adoption of AI-assisted methodologies to enhance clinical workflow.