A method for incorporating organ motion due to breathing into 3D dose calculations in the liver: sensitivity to variations in motion

• Published in: Medical physics (Lancaster) Vol. 30; no. 10; p. 2643
• Main Authors: Lujan, Anthony E, Balter, James M, Ten Haken, Randall K
• Format: Journal Article
• Language: English
• Published: United States 01.10.2003
• Subjects: Algorithms; Clinical Trials as Topic; Humans; Liver - diagnostic imaging; Liver Neoplasms - radiotherapy; Models, Statistical; Motion; Radiography; Radiometry; Radiotherapy Planning, Computer-Assisted - methods; Respiration
• ISSN: 0094-2405
• DOI: 10.1118/1.1609057

Organ motion has been previously described using a probability distribution function that depends solely upon the amplitude of motion and the degree of asymmetry in the breathing cycle, and that function has been used with patient specific parameters to correct static dose distributions for patient breathing using a dose convolution method. In this study, the consequences of errors in the selection of those two parameters were evaluated. Patients previously treated using a focal liver dose escalation protocol were selected with tumors located in the superior or inferior portion of the liver. For a fixed degree of asymmetry (amplitude), the amplitude (asymmetry) of motion was varied about its nominal value and the consequences of organ motion on the dose distribution and the (potentially new) prescription dose were evaluated. These comparisons show that small (+/- 3 mm) variations of the amplitude of motion about the nominally measured value may not result in clinically significant changes (< a single fraction change in the prescription dose), however, larger variations (> 5 mm) can lead to significant changes. Assuming from measurement that the patient breathes asymmetrically (spends more time at expiration), variations in the assumed degree of asymmetry rarely lead to clinically significant changes; the most significant cause for concern being when the patient breathing cycle is maximally different from the treatment planning case (e.g., patient assumed to spend more time at expiration, but later breaths symmetrically). The results point out where quality assurance efforts should be concentrated to help assure the validity of the assumptions used to correct the static dose distributions for patient breathing using the convolution method.