Accelerating CEST imaging using a model‐based deep neural network with synthetic training data

• Published in: Magnetic resonance in medicine Vol. 91; no. 2; pp. 583 - 599
• Main Authors: Xu, Jianping, Zu, Tao, Hsu, Yi‐Cheng, Wang, Xiaoli, Chan, Kannie W. Y., Zhang, Yi
• Format: Journal Article
• Language: English
• Published: Hoboken Wiley Subscription Services, Inc 01.02.2024
• Subjects: Ablation; Artificial neural networks; Brain; Brain cancer; Brain mapping; Brain tumors; Coils; Deep learning; Image processing; Image quality; Image reconstruction; Machine learning; Medical imaging; Neural networks; Neuroimaging; Protons; Redundancy; Tumors
• ISSN: 0740-3194; 1522-2594; 1522-2594
• DOI: 10.1002/mrm.29889

PurposeTo develop a model‐based deep neural network for high‐quality image reconstruction of undersampled multi‐coil CEST data.Theory and MethodsInspired by the variational network (VN), the CEST image reconstruction equation is unrolled into a deep neural network (CEST‐VN) with a k‐space data‐sharing block that takes advantage of the inherent redundancy in adjacent CEST frames and 3D spatial–frequential convolution kernels that exploit correlations in the x‐ω domain. Additionally, a new pipeline based on multiple‐pool Bloch–McConnell simulations is devised to synthesize multi‐coil CEST data from publicly available anatomical MRI data. The proposed network is trained on simulated data with a CEST‐specific loss function that jointly measures the structural and CEST contrast. The performance of CEST‐VN was evaluated on four healthy volunteers and five brain tumor patients using retrospectively or prospectively undersampled data with various acceleration factors, and then compared with other conventional and state‐of‐the‐art reconstruction methods.ResultsThe proposed CEST‐VN method generated high‐quality CEST source images and amide proton transfer‐weighted maps in healthy and brain tumor subjects, consistently outperforming GRAPPA, blind compressed sensing, and the original VN. With the acceleration factors increasing from 3 to 6, CEST‐VN with the same hyperparameters yielded similar and accurate reconstruction without apparent loss of details or increase of artifacts. The ablation studies confirmed the effectiveness of the CEST‐specific loss function and data‐sharing block used.ConclusionsThe proposed CEST‐VN method can offer high‐quality CEST source images and amide proton transfer‐weighted maps from highly undersampled multi‐coil data by integrating the deep learning prior and multi‐coil sensitivity encoding model.