Direct integration of deep learning-based GTV auto-segmentation into a clinical radiotherapy planning system

Manual Gross Tumor Volume (GTV) delineation for radiotherapy is time‑consuming and prone to inter‑observer variability. Deep Learning (DL)-based auto‑segmentation offers faster and more consistent solutions, but translation into clinical practice remains limited. This study aimed to bridge this gap by integrating an in‑house GTV auto‑segmentation DL model into a radiotherapy Treatment Planning System (TPS). To demonstrate the integration process, a head‑and‑neck GTV segmentation model was successfully deployed and operated seamlessly within the TPS. In this integration, DL‑generated GTV contours were produced in under three minutes, demonstrating the feasibility of DL‑based GTV segmentation in the TPS without disrupting the radiotherapy workflow.