Publication: Study on voxel-grade pulmonary function imaging based on ventilation-perfusion mapping and its application in radiotherapy
No Thumbnail Available
Date
2025-12
Authors
Suyan, Bi
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Current single-modality lung functional imaging techniques for lung cancer radiotherapy (RT) fail to comprehensively reflect the spatial distribution characteristics of pulmonary ventilation and perfusion functions，hindering precise radiotherapy (RT) planning for lung cancer. This study aimed to develop and evaluate a novel voxel-level ventilation-perfusion (VP) imaging technique for lung cancer RT optimization. A retrospective analysis was performed on 20 lung cancer patients who underwent four-dimensional computed tomography (4D-CT) and positron emission tomography/computed tomography (PET/CT) scans. ventilation (V-Imaging) was derived from 4D-CT deformable registration, and perfusion (P-Imaging) from pre-treatment PET/CT. An integrated VP-Imaging was generated via Hadamard product. Agreement between modalities was assessed using Dice Similarity Coefficient (DSC) and correlation. Intensity-modulated RT (IMRT) plans were optimized based on each functional image, comparing doses to targets and organs at risk. VP-Imaging demonstrated superior concordance with single-modality images (DSC up to 0.71, Pearson CC up to 0.943) compared to the poor agreement between V- and P-Imaging alone (DSC=0.11 in high-function regions). VP-guided plans yielded significantly different dose distributions (p<0.05) in high-function ipsilateral lung compared to V- or P-based plans, without compromising other OAR doses. In conclusion, VP-Imaging successfully integrates ventilation and perfusion data from routine clinical scans, resolving V/Q mismatch and providing a superior map of functional lung. It facilitates personalized RT planning by enabling selective sparing of high-function tissue, potentially reducing pulmonary toxicity while maintaining treatment efficacy.