SU-E-J-82: Improvement in Reproducibility of Lung Expansion Measures with Respiratory Effort Correction

Longitudinal measurements of pulmonary function must account for subject variation when assessing radiation-induced changes. Previously, we reported intra-subject reproducibility of Jacobian-based measures of lung tissue expansion using repeat 4DCT scans prior to radiation therapy without correcting for differences in respiratory effort. In this study, we present two normalization schemes that correct ventilation images for variations in respiratory effort.

Two repeat 4DCT image acquisitions were collected before treatment from seven patients. Using a tissue volume preserving deformable image registration algorithm, two Jacobian ventilation maps were computed from separate acquisitions. Two effort normalization strategies were investigated using intermediated inspiration phases upon the principles of equivalent tidal volume (ETV) and equivalent lung volume (ELV). Additional experiments were performed to verify the effectiveness of the ETV method. Scatter plots of two ventilation maps and statistical parameters of Jacobian ratio were compared before and after each effort correction approach.

The scatter plots of the ventilation maps show improvement in measurement reproducibility using either the ETV or ELV effort correction for all but one case. The one case that did not show improvement had a very similar lung volumes and tidal volumes in the two studies, so effort correction was unnecessary. The coefficient of variation (CV) of Jacobian ratio improved by 21±11 percent (mean±standard deviation) after ETV, and improved by 20±17 percent after ELV. Our experimental results on ETV show the decrease of tidal volume difference in two acquisitions improves reproducibility of lung expansion measures.

ETV and ELV correction methods improve the reproducibility of pulmonary function measurements in subjects that have differences in respiratory effort in the baseline and followup scans. Additional work is needed to investigate whether ETV or ELV is more effective, and to develop alternate regional normalization schemes that can account for difference in lung expansion rates.