AbstractBackground: Head and neck cancers (HNC) are complex to treat due to their proximity to vital structures. Accurate tumor delineation is essential for effective radiotherapy, with 18F-FDG PET/CT offering functional insights to potentially enhance targeting accuracy over CT alone. This study aims to compare gross tumor volumes (GTV) from CT and PET/CT-based methods, evaluating adaptive and fixed threshold segmentation techniques.
Materials and Methods: This Eight Months study included 10 patients with biopsy-proven HNC who underwent IMRT treatment planning using CT and PET/CT imaging. Tumor volumes were contoured on CT scans and PET/CT images using the SBR adaptive threshold method and fixed SUV thresholds at 40% and 50%. The delineated volumes were compared for differences in tumor size and accuracy.
Results: PET-based adaptive thresholding yielded smaller GTVs than CT, with significant differences observed between SBR and fixed thresholds. The fixed SUV 40% and 50% methods often failed to delineate tumors fully in certain cases, highlighting adaptive methods as superior for accurate tumor contouring. The adaptive technique provided the most reliable volume estimates, particularly for nodal GTVs.
Conclusion: PET/CT with adaptive thresholding is a superior method for tumor delineation in HNC, potentially enhancing radiotherapy precision and treatment outcomes. Fixed threshold methods, while helpful, are less consistent and may fail in complex cases.
