Chronic obstructive pulmonary disease (COPD) is a highly heterogeneous disease with various phenotypes. Registered inspiratory and expiratory CT images can generate the parametric response map (PRM) that characterizes phenotypes' spatial distribution and proportions. However, increased radiation dosage, scan time, quality control, and patient cooperation requirements limit the utility of PRM. This study aims to synthesize a PRM using only inspiratory CT scans. First, a CycleGAN with perceptual loss and a multiscale discriminator (MPCycleGAN) is proposed and trained to synthesize registered expiratory CT images from inspiratory images. Next, a strategy named InspirationOnly is introduced, where synthesized images replace actual expiratory CT images. The image synthesizer outperformed state-of-the-art models, achieving a mean absolute error of 105.66 ± 36.64 HU, a peak signal-to-noise ratio of 21.43 ± 1.87 dB, and a structural similarity of 0.84 ± 0.02. The intraclass correlation coefficients of emphysema, fSAD, and normal proportions between the InspirationOnly and ground truth were 0.995, 0.829, and 0.914, respectively. The proposed MPCycleGAN enables the InspirationOnly strategy to yield PRM using only inspiratory CT. The estimated COPD phenotypes are consistent with those from dual-phase CT and correlated with the spirometry parameters. This offers a potential tool for characterizing phenotypes of COPD, particularly when expiratory CT images are unavailable.