Open AccessImpact of different image guidance schedules in head and neck irradiation: A retrospective analysis
Author(s) -
Bhatia Jatin Vinod,
Kothavade Vikas Digambar,
Ramachandran Vinoth Kumar
Publication year - 2021
Publication title -
precision radiation oncology
Language(s) - English
Resource type - Journals
ISSN - 2398-7324
DOI - 10.1002/pro6.1109
Subject(s) - radiation therapy , head and neck cancer , margin (machine learning) , radiation treatment planning , cone beam computed tomography , nuclear medicine , head and neck , medicine , random error , computed tomography , image registration , cone beam ct , schedule , computer science , mathematics , image (mathematics) , radiology , surgery , statistics , artificial intelligence , operating system , machine learning
Objective Is daily image‐guided radiation therapy justified for all patients? Here, we attempt to address this issue and present data obtained from various frequencies. Methods All consecutive head and neck cancer patients receiving radiation with curative intent and daily image‐guided radiation therapy using kV cone beam computed tomography scans were eligible. Displacements in the medial–lateral, superior–inferior (SI), and anterior–posterior directions were documented daily. The retrospectively simulated scenarios with reducing frequency were 50%, 30%, 20%, and 10%. The mean displacements, systematic errors, random errors, and planning target volume margins were calculated. Results A total of 831 cone beam computed tomography scans of 31 patients were analyzed, showing normally distributed mean displacements in the medial–lateral, SI, and anterior–posterior direction, except in SI for daily format. The mean displacements in the three directions was <0.1 cm, reflected in the systematic errors (<0.2 cm). In the SI direction, the random errors for daily format were 0.7 cm and 0.2 cm for simulated scenarios; these differences were demonstrated in planning target volume margin calculation, 1 cm in the SI direction based on daily data, and 0.5 cm based on simulated scenarios. The margin estimation was similar in the other directions for all formats. Conclusion The errors were similar across all scenarios, inconclusive of superiority for any schedule; however, a higher random error was observed with daily format in the SI direction. Our findings further validate the need for more data to determine the optimum image‐guided radiation therapy frequency.