Abstract
To improve the image quality of highly accelerated multi-channel MRI data by learning a joint variational network that reconstructs multiple clinical contrasts jointly. Data from our multi-contrast acquisition were embedded into the variational network architecture where shared anatomical information is exchanged by mixing the input contrasts. Complementary k-space sampling across imaging contrasts and Bunch-Phase/Wave-Encoding were used for data acquisition to improve the reconstruction at high accelerations. At 3T, our joint variational network approach across T1w, T2w and T2-FLAIR-weighted brain scans was tested for retrospective under-sampling atR=6 (2D) andR=4×4 (3D) acceleration. Prospective acceleration was also performed for 3D data where the combined acquisition time for whole brain coverage at 1mm isotropic resolution across three contrasts was less than 3 min. Across all test datasets, our joint multi-contrast network better preserved fine anatomical details with reduced image-blurring when compared to the corresponding single-contrast reconstructions. Improvement in image quality was also obtained through complementary k-space sampling and Bunch-Phase/Wave-Encoding where the synergistic combination yielded the overall best performance as evidenced by exemplary slices and quantitative error metrics. By leveraging shared anatomical structures across the jointly reconstructed scans, our joint multi-contrast approach learnt more efficient regularizers, which helped to retain natural image appearance and avoid over-smoothing. When synergistically combined with advanced encoding techniques, the performance was further improved, enabling up to R=16-fold acceleration with good image quality. This should help pave the way to very rapid high-resolution brain exams.
Highlights
Fast imaging techniques have been widely adopted into clinical practice to speed up MRI scans and, help improve patient throughput, reduce the sensitivity to involuntary patient motion,[1] improve patient compliance, and potentially obviate the need for sedation in pediatric patients.[2]
The single-contrast variational network (VN) mitigated most of these issues, but the artifact and noise reduction came at the cost of over-smoothing and loss of spatial resolution, as indicated by the zoom-in
Fine anatomical details were better preserved as demonstrated by the improved conspicuity of a region of cerebrospinal fluid (CSF) in the posterior of the brain for VN + Bunch Phase Encoding (BPE)
Summary
Fast imaging techniques have been widely adopted into clinical practice to speed up MRI scans and, help improve patient throughput, reduce the sensitivity to involuntary patient motion,[1] improve patient compliance, and potentially obviate the need for sedation in pediatric patients.[2]. Inspired by traditional iterative techniques for inverse problems, several approaches[29,30,31] have posed the MRI image reconstruction as an unrolled gradient descent optimization where the physics model is embedded in the reconstruction and regularizers/priors are learnt from training data This formulation can be understood as a generalization of CS where neural networks are utilized instead of hand-crafted domain transformations (such as wavelet or TV). With this framework many existing physics- and CS-based techniques have been outperformed while enabling much shorter reconstruction times.[29] In a recent work[32] such a network was utilized to reconstruct a highly accelerated Wave acquisition where imperfections of the sinusoidal Wave gradient trajectory were automatically estimated by the network without additional time-consuming optimizations (eg AutoPSF33).
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
