Characterization Of Breathing Patterns Of Patients Undergoing Respiratory-Correlated Imaging

Purpose/Objective: Respiratory-induced tumor motion limits the effectiveness of radiotherapy for thoracic lesions. To account for respiratory-induced intrafraction motion an internal target volume (ITV) that includes tumor excursion during the respiratory cycle can be defined. We currently determine ITVs in two ways: from respiratory-correlated 4DCT images and from feed-back guided breath-hold (FGBH) images taken at end-inspiration and at end-expiration. A major assumption of the FGBH approach is that inspiration and expiration breath-hold GTVs accurately represent the limits of the free-breathing (FB) GTV motion. We performed a preliminary test of this assumption on a sample of 12 patients. Materials/Methods: We currently have respiratory traces (Varian RPMTM) for over 150 patients that have undergone both FGBH and 4DCT procedures with concurrent recordings of their respiratory motion. We created a software-based quality assurance (QA) tool in MATLAB for statistically analyzing patient breathing patterns during respiratory-correlated imaging. From the FGBH traces we computed the mean position of the marker during inspiration breath-hold (BHinsp) and during expiration breath-hold (BHexp) and their standard deviations (SD). From the FB traces, we computed equivalent values: mean end-inspiration position (FBinsp) and end-expiration position (FBexp) and their SDs. We also computed the ranges of motion during individual breath-holds and the variations in breathing cycle amplitudes to evaluate their stability. BHinsp and BHexp were compared to FBinsp and FBexp, respectively, using a student′s t-test. The ratio of the breath-hold range of motion to free-breathing cycle amplitude variation was also calculated as a measure of patient compliance during BH. Results: Analysis of FB traces revealed that the standard deviation of EI values was, on average, 2.3 times larger compared to EE values (larger in 11 cases (92 %). Breath-hold at inspiration (BHinsp) was significantly different (p < 0.05) from the mean free breathing position at EI in 8 of 12 patients (66 %). Breath-hold at expiration (BHexp) was significantly different (p < 0.05) from the mean free breathing position at EE in 11 of 12 patients (92 %). The range of motion during BHinsp was larger (i.e. a less stable BH) than at BHexp in 11 of 12 patients (92 %) and also larger when expressed as a percentage of the mean FB amplitude. Conclusions: We have demonstrated the use of a QA tool for rapidly analyzing respiratory traces acquired from patients undergoing respiratory-correlated imaging for the purpose of respiratory correlated therapy. Currently it is being used for retrospective analysis, but in the future it will be used to analyze traces as they are acquired to immediately determine clinical acceptability. In this small sampling of patients, the analysis of FB traces revealed more reproducible position at end-expiration for all patients studied. This indicated that end-expiration as opposed to end-inspiration is the best phase of the respiratory cycle for passive respiratory-gated treatments. Analyses of the range of motion during BH as a percentage of the mean FB amplitude reveal that patients achieved more stable breath-hold at end-expiration than end-inspiration. Hence for breath-hold treatment protocols, breath-hold at the end expiration position would be prudent. When BH positions were compared to FB traces, it was found that the FGBH ITV did not always accurately represent the limits of the FB GTV motion, especially at expiration. Therefore, either great care must be taken during BH imaging or a different approach, such as 4DCT, should be used to determine the ITV.