ESTRO 2025 - Abstract Book

S2561

Physics - Autosegmentation

ESTRO 2025

Keywords: autosegmentation, organizational impact

References: [1] K. Mackay, D. Bernstein, B. Glocker, K. Kamnitsas, A. Taylor,A Review of the Metrics Used to Assess Auto-Contouring Systems in Radiotherapy, Clinical Oncology, 2023 [2] Radici, L.; Ferrario, S.; Borca, V.C.; Cante, D.; Paolini, M.; Piva, C.; Baratto, L.; Franco, P.; La Porta, M.R. Implementation of a Commercial Deep Learning-Based Auto Segmentation Software in Radiotherapy: Evaluation of Effectiveness and Impact on Workflow. Life, 2022 [3] Patil, Mangesh and Mulinti, Suneetha and Aggarwal, Sumeet and Singamsetty, Multi-Institutional Study to Evaluate an AI Algorithm for Auto-Segmentation of the Head and Neck Nodal Contouring, AI in Precision Oncology, 2024

4284

Digital Poster Future liver remnant meets the future of medicine: AI integration in liver metastases assessment and treatment selection Sabina E Sucuri 1,2 , Maria Timpea 3,2 , Alexandru V Zariosu 3,2 , Bogdan Chivu 4,2 , Pop C Cristina 4,2 , Bogdan Mocanu 1,2 , Marius Stanescu 5 , Lucian Bicsi 5 , Dragos Duse 5 , Dragos Grama 5 , Remus C Stoica 6,2 1 Radiation Oncology, Coltea Clinical Hospital, Bucharest, Romania. 2 Medical Department, Synaptiq, Cluj-Napoca, Romania. 3 Radiation Oncology, Institute of Oncology Prof. Dr. Alexandru Trestioreanu, Bucharest, Romania. 4 Radiation Oncology, Neolife, Bucharest, Romania. 5 Research Department, Synaptiq, Cluj-Napoca, Romania. 6 Radiation Oncology, Global Medical Health, Bucharest, Romania Purpose/Objective: Surgical resection is the standard of care for liver metastases due to its curative potential. However, it is not feasible for patients with insufficient future liver remnant (FLR) or poor liver function. SBRT has emerged as a non-invasive alternative, offering high local control rates while preserving liver function. This study evaluates the integration of AI-powered auto-segmentation and prompt-based auto-contouring tools into radiation oncology workflows to assist in decision-making between SBRT and surgery by calculating the remaining liver volume: Liver-GTV. Material/Methods: This proof-of-concept study included 10 patients with liver metastases. Imaging protocols used native and contrast enhanced CT scans for tumor board evaluations. AI-driven tools integrated within the Mediq TPS from Synaptiq streamlined auto-contouring. The liver was auto-segmented to define boundaries, and GTV(s) were delineated using a prompt-based tool, automating tumor contouring across all slices. FLR was calculated by subtracting the GTV from the liver volume, ensuring it met safety thresholds: • For a normal liver function: FLR ≥20-30% (~700–800 cc). • For an impaired liver function: FLR ≥40% (~1,000 cc). Key parameters analyzed included AI tools' time and iteration count for contouring and feasibility of surgical resection based on FLR assessment, helping multidisciplinary teams select optimal treatment.