Anatomy-wise lung ventilation imaging for precise functional lung avoidance radiation therapy.

Abstract

Objective.This study aimed to propose a method for obtaining anatomy-wise lung ventilation image (VI_aw) that enables functional assessment of lung parenchyma and tumor-blocked pulmonary segments. The VI_awwas used to define multiple functional volumes of the lung and thereby support radiation treatment planning.Approach.A super-voxel-based method was employed for functional assessment of lung parenchyma to generate VI_svd. In the VI_svdof the 11 patients with tumor blockage of the airway, the functional value in tumor-blocked segments was set to 0 to generate the VI_aw. The lung was divided into regions of high functional volume (HFV), unrecoverable low functional volume (LFV), and recoverable LFV (rLFV, the region in the tumor-blocked segment with a high function value based on the VI_svd) to design three intensity-modulated photon plans for five patients. These plans were an anatomical-lung-guided plan (aPlan), a functional-lung-guided plan (fPlan), and a recoverable functional-lung-guided plan (rfPlan) where the latter protected both HFV and rLFV.Main results.The LFV in the reference ventilation images and the tumor-blocked segments had a high overlap similarity coefficient value of 0.90 ± 0.07. The mean Spearman correlation between the VI_awand reference ventilation images was 0.72 ± 0.05 for the patient with tumor blockage of the airway. TheV20 and mean dose of rLFV in rfPlan were lower than those in aPlan by 12.1 ± 8.4% and 13.0 ± 6.4%, respectively, and lower than those in fPlan by 14.9 ± 9.8% and 15.9 ± 6.5%, respectively.Significance.The VI_awcan reach a moderate-strong correlation with reference ventilation images and thus can identify rLFV to support treatment planning to preserve lung function.