SU‐E‐J‐235: Audiovisual Biofeedback Improves the Correlation Between Internal and External Respiratory Motion

Abstract

Purpose:External respiratory surrogates are often used to predict internal lung tumor motion for beam gating but the assumption of correlation between external and internal surrogates is not always verified resulting in amplitude mismatch and time shift. To test the hypothesis that audiovisual (AV) biofeedback improves the correlation between internal and external respiratory motion, in order to improve the accuracy of respiratory‐gated treatments for lung cancer radiotherapy.Methods:In nine lung cancer patients, 2D coronal and sagittal cine‐MR images were acquired across two MRI sessions (pre‐ and mid‐treatment) with (1) free breathing (FB) and (2) AV biofeedback. External anterior‐posterior (AP) respiratory motions of (a) chest and (b) abdomen were simultaneously acquired with physiological measurement unit (PMU, 3T Skyra, Siemens Healthcare Erlangen, Germany) and real‐time position management (RPM) system (Varian, Palo Alto, USA), respectively. Internal superior‐inferior (SI) respiratory motions of (c) lung tumor (i.e. centroid of auto‐segmented lung tumor) and (d) diaphragm (i.e. upper liver dome) were measured from individual cine‐MR images across 32 dataset. The four respiratory motions were then synchronized with the cine‐MR image acquisition time. Correlation coefficients were calculated in the time variation of two nominated respiratory motions: (1) chest‐abdomen, (2) abdomen‐diaphragm and (3) diaphragm‐lung tumor. The three combinations were compared between FB and AV biofeedback.Results:Compared to FB, AV biofeedback improved chest‐abdomen correlation by 17% (p=0.005) from 0.75±0.23 to 0.90±0.05 and abdomen‐diaphragm correlation by 4% (p=0.058) from 0.91±0.11 to 0.95±0.05. Compared to FB, AV biofeedback improved diaphragm‐lung tumor correlation by 12% (p=0.023) from 0.65±0.21 to 0.74±0.16.Conclusions:Our results demonstrated that AV biofeedback significantly improved the correlation of internal and external respiratory motion, thus suggesting the need of AV biofeedback in respiratory‐gated treatments.