AUTHOR=Liu Zhiqiang , Tian Yuan , Miao Junjie , Men Kuo , Wang Wenqing , Wang Xin , Zhang Tao , Bi Nan , Dai Jianrong 

TITLE=Deriving Pulmonary Ventilation Images From Clinical 4D-CBCT Using a Deep Learning-Based Model

JOURNAL=Frontiers in Oncology

VOLUME=Volume 12 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2022.889266

DOI=10.3389/fonc.2022.889266

ISSN=2234-943X

ABSTRACT=Purpose: The current algorithms for measuring ventilation images from 4D cone-beam computed tomography (CBCT) are affected by the accuracy of deformable image registration (DIR). This study proposes new deep learning (DL) method that does not rely on DIR to derive ventilation image from 4D-CBCT (CBCT-VI), which was validated with the gold standard single-photon emission-computed tomography ventilation image (SPECT-VI). 

Materials and Method: This study consists of 4D-CBCT and 99mTc-Technegas SPECT/CT scans of 28 esophagus or lung cancer patients. The scans were rigidly registered for each patient. Using these data, the CBCT-VI was derived using a deep-learning-based model. Two types of model input data are studied, namely, (a) ten phases of 4D-CBCT and (b) two phases of peak-exhalation and peak-inhalation of 4D-CBCT. A sevenfold cross-validation was applied to train and evaluate the model. The DIR-dependent methods (density-change-based and Jacobian-based) were used to measure the CBCT-VIs for comparison. The correlation was calculated between each CBCT-VI and SPECT-VI using voxel-wise Spearman’s correlation. The ventilation images were divided into high, medium, and low functional lung (HFL, MFL, and LFL) regions. The similarity of different functional lung between SPECT-VI and each CBCT-VI was evaluated using the dice similarity coefficient (DSC). One-factor ANONA model was used for statistical analysis of averaged DSC for different methods of generating ventilation images.

Results: The correlation values were (0.02±0.10), (0.02±0.09), and (0.65±0.13)/ (0.65±0.15), and the averaged DSC values were (0.34±0.04), (0.34±0.03) and (0.59±0.08)/(0.58±0.09) for the density-change, Jacobian and deep learning methods, respectively. The strongest correlation and the highest similarity with SPECT-VI were observed for the deep learning method compared to the density-change and Jacobian methods.

Conclusion: The results showed that the deep learning method improved the accuracy of correlation and similarity significantly, and the derived CBCT-VIs have the potential to monitor the lung function dynamic changes during radiotherapy.