The RadiomicART project aims to harness artificial intelligence to enhance treatment personalization for patients with head and neck squamous cell carcinoma (HNSCC) undergoing Adaptive Radiotherapy (ART). Traditional radiotherapy, while effective, carries significant risks of locoregional failure and radiation-induced toxicity, such as xerostomia. ART leverages advanced imaging technologies, including CT, MRI, and PET-FDG, to dynamically adjust treatment plans based on patient-specific anatomical changes during the therapy course.
This project focuses on developing machine learning models that integrate radiomic features—quantitative biomarkers extracted from medical images—with clinical and dosimetric data. By analyzing imaging data collected at multiple time points (before abd during radiotherapy), the models aim to predict patient outcomes and the risk of developing adverse toxicities, such as damage to the salivary glands.
The framework developed was employed on the prediction of xerostomia at 3 and 6 months post-treatment by identifying key features from organs at risk, including the parotid and submandibular glands. This approach not only informs treatment planning but also promotes more organ-sparing strategies, potentially reducing long-term toxicity and improving patient quality of life.
RadiomicART represents a significant step toward personalized cancer therapy, demonstrating how AI-driven insights can optimize radiotherapy for better outcomes and reduced side effects.