Abstract: Objective: To explore the feasibility of using dual low-dose computed tomography (CT) combined with a deep learning image reconstruction (DLIR) algorithm to optimize pulmonary artery mixed reality (MR) imaging. Methods: One-hundred-and-eight patients who planned to undergo pulmonary artery MR imaging at our hospital were prospectively enrolled into this study. They were randomly divided into a low-dose pulmonary artery MR group (Group A: low-dose CT combined with DLIR algorithm, using tube current of 70 kVp and contrast agent of 300 mgI/mL, n=53) and a conventional-dose pulmonary artery MR group (Group B: conventional-dose CT, using tube voltage of 120 kVp and contrast agent of 370 mgI/mL, n=55). In both groups A and B, automatic tube current modulation was applied to the tube currents. Subsequently, the CT DICOM data obtained from the two groups were imported into three-dimensional (3D) visualization modelling software to complete the MR 3D modelling. Demographic characteristics and radiation dose-related indicators of the two patient groups were compared using an independent sample t-test. Additionally, Pearson and Bland-Altman analyses were used to evaluate the consistency of pulmonary artery MR between the two groups, whereas the subjective evaluation of MR quality was conducted using the Mann-Whitney U test. Furthermore, the consistency of the evaluating factors was analyzed using Kappa analysis. Results: Compared to Group B, the radiation dose in Group A decreased by approximately 41%, whereas the quality of the CT DICOM data gradually improved with increasing DLIR intensity. The DICOM images obtained using high-strength DLIR (DLIR-H) demonstrated the best quality. However, there was no significant difference in the quality of the MR between groups A and B. In the Bland-Altman analysis, there was good consistency between Groups A and Group B. Moreover, the subjective consistency of the evaluating factors was good, with a Kappa value of 0.76. Conclusion: The combination of dual low-dose CT and the DLIR algorithm can effectively reduce the radiation and contrast agent doses of pulmonary artery MR without affecting the quality of pulmonary artery MR imaging, laying the foundation for its clinical application and future advancement.