Deep Learning Single-Frame and Multiframe Super-Resolution for Cardiac MRI

Background: Cardiac MRI is limited by long acquisition times, yet faster acquisition of smaller-matrix images reduces spatial detail. Deep learning (DL) might enable both faster acquisition and higher spatial detail via super-resolution. Purpose: To explore the feasibility of using DI. to enhance spatial detail from small-matrix MRI acquisitions and evaluate its performance against that of conventional image upscaling methods. Materials and methods: Short-axis cite cardiac MRI examinations performed between January 2012 and December 2018 at one institution were retrospectively collected for algorithm development and testing. Convolutional neural networks (CNNs), a form of DNNs were trained to perform super resolution in image space by using as generated low-resolution data. There were 7090, 2090, and 1090 of examinations allocated to training, validation, and test sets, respectively. CNNs were compare' against bicubic interpolation and Fourier-based zero padding by calculating the structural similarity index (SSIM) between high-resolution ground truth and each upscaling method. Means and standard deviations of the SSIM were reported, and statistical significance was determined by using the Wilcoxon signed-rank test. For evaluation of clinical performance, left ventricular volumes were measured, and statistical significance was determine by using the paired Student t test. Results: For CNN training and retrospective analysis, 400 MRI scans front 367 patients (mean age, 48 years +/- 18; 214 men) were included. All CNNs outperformed zero padding and bicubic interpolation at upsatnpling factors from two to 64 (P < .001). CNNs outperformed zero padding on more than 90.2% of sliccs (9828 of 9907). In addition, 10 patients (mein age, 51 +/- years 22; seven men) were perspective recruied for super-resolution MRI. Super-resolved low-resolution images yielded left ventricular volumes comparable to those from full-resolution images (P > .05), and super-resolved full-resolution images appeared to further enhance anatomic detail. Conclusion: Deep learning outperformed conventional upscaling methods and recovered high-frequency spatial information. Although training was performed only on short-axis cardiac MRI examinations, the proposed strategy appeared to improve quality in other imaging planes. (C) RSNA, 2020