Multi-input cardiac image super-resolution using convolutional neural networks

Oktay O, Bai W, Lee M, Guerrero R, Kamnitsas K, Caballero J, De Marvao A, Cook S, O’Regan D, Rueckert D (2016)

Publication Type: Conference contribution

Publication year: 2016

Journal

Lecture Notes in Computer Science Springer Verlag

Publisher: Springer Verlag

Book Volume: 9902 LNCS

Pages Range: 246-254

Conference Proceedings Title: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

ISBN: 9783319467252

DOI: 10.1007/978-3-319-46726-9_29

Abstract

3D cardiac MR imaging enables accurate analysis of cardiac morphology and physiology. However,due to the requirements for long acquisition and breath-hold,the clinical routine is still dominated by multi-slice 2D imaging,which hamper the visualization of anatomy and quantitative measurements as relatively thick slices are acquired. As a solution,we propose a novel image super-resolution (SR) approach that is based on a residual convolutional neural network (CNN) model. It reconstructs high resolution 3D volumes from 2D image stacks for more accurate image analysis. The proposed model allows the use of multiple input data acquired from different viewing planes for improved performance. Experimental results on 1233 cardiac short and long-axis MR image stacks show that the CNN model outperforms state-of-the-art SR methods in terms of image quality while being computationally efficient. Also,we show that image segmentation and motion tracking benefits more from SR-CNN when it is used as an initial upscaling method than conventional interpolation methods for the subsequent analysis.

Involved external institutions

Imperial College London / The Imperial College of Science, Technology and Medicine GB United Kingdom (GB) Magic Pony Technology Ltd. GB United Kingdom (GB)

How to cite

APA:

Oktay, O., Bai, W., Lee, M., Guerrero, R., Kamnitsas, K., Caballero, J.,... Rueckert, D. (2016). Multi-input cardiac image super-resolution using convolutional neural networks. In Leo Joskowicz, Mert R. Sabuncu, William Wells, Gozde Unal, Sebastian Ourselin (Eds.), Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (pp. 246-254). Springer Verlag.

MLA:

Oktay, Ozan, et al. "Multi-input cardiac image super-resolution using convolutional neural networks." Proceedings of the Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) Ed. Leo Joskowicz, Mert R. Sabuncu, William Wells, Gozde Unal, Sebastian Ourselin, Springer Verlag, 2016. 246-254.

BibTeX: Download