Ze-Peng Ma,1,2,* Yue-Ming Zhu,3,* Xiao-Dan Zhang,4 Yong-Xia Zhao,1 Wei Zheng,3 Shuang-Rui Yuan,1 Gao-Yang Li,1 Tian-Le Zhang1 

1Department of Radiology, Affiliated Hospital of Hebei University/ Clinical Medical College, Hebei University, Baoding, 071000, People’s Republic of China; 2Hebei Key Laboratory of Precise Imaging of inflammation Tumors, Baoding, Hebei Province, 071000, People’s Republic of China; 3College of Electronic and Information Engineering, Hebei University, Baoding, Hebei Province, 071002, People’s Republic of China; 4Department of Ultrasound, Affiliated Hospital of Hebei University, Baoding, Hebei Province, 071000, People’s Republic of China

*These authors contributed equally to this work

Correspondence: Xiao-Dan Zhang, Department of Ultrasound, Affiliated Hospital of Hebei University, No. 212 of Yuhua East Road, Lianchi District, Baoding, 071000, People’s Republic of China, Tel +86 17325535302, Email xiaodanzhangzxd@126.com

Objective: To evaluate the effectiveness of deep learning technology based on generative adversarial networks (GANs) in reducing motion artifacts in cardiac magnetic resonance (CMR) cine sequences.
Methods: The training and testing datasets consisted of 2000 and 200 pairs of clear and blurry images, respectively, acquired through simulated motion artifacts in CMR cine sequences. These datasets were used to establish and train a deep learning GAN model. To assess the efficacy of the deep learning network in mitigating motion artifacts, 100 images with simulated motion artifacts and 37 images with real-world motion artifacts encountered in clinical practice were selected. Image quality pre- and post-optimization was assessed using metrics including Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM), Leningrad Focus Measure, and a 5-point Likert scale.
Results: After GAN optimization, notable improvements were observed in the PSNR, SSIM, and focus measure metrics for the 100 images with simulated artifacts. These metrics increased from initial values of 23.85± 2.85, 0.71± 0.08, and 4.56± 0.67, respectively, to 27.91± 1.74, 0.83± 0.05, and 7.74± 0.39 post-optimization. Additionally, the subjective assessment scores significantly improved from 2.44± 1.08 to 4.44± 0.66 (P< 0.001). For the 37 images with real-world artifacts, the Tenengrad Focus Measure showed a significant enhancement, rising from 6.06± 0.91 to 10.13± 0.48 after artifact removal. Subjective ratings also increased from 3.03± 0.73 to 3.73± 0.87 (P< 0.001).
Conclusion: GAN-based deep learning technology effectively reduces motion artifacts present in CMR cine images, demonstrating significant potential for clinical application in optimizing CMR motion artifact management.

Keywords: cardiac magnetic resonance, deep learning, generative adversarial networks, image quality, motion artifacts