J-MoDL: Joint Model-Based Deep Learning for Optimized Sampling and Reconstruction

Hemant Kumar Aggarwal; Mathews Jacob

doi:10.1109/JSTSP.2020.3004094

Back

Journal article

J-MoDL: Joint Model-Based Deep Learning for Optimized Sampling and Reconstruction

Hemant Kumar Aggarwal and Mathews Jacob

IEEE journal of selected topics in signal processing, Vol.14(6), pp.1151-1162

10/2020

DOI: 10.1109/JSTSP.2020.3004094

PMCID: PMC7893809

PMID: 33613806

Files and links (1)

url

https://arxiv.org/pdf/1911.02945View

Open Access

Abstract

Modern MRI schemes, which rely on compressed sensing or deep learning algorithms to recover MRI data from undersampled multichannel Fourier measurements, are widely used to reduce the scan time. The image quality of these approaches is heavily dependent on the sampling pattern. In this article, we introduce a continuous strategy to optimize the sampling pattern and the network parameters jointly. We use a multichannel forward model, consisting of a non-uniform Fourier transform with continuously defined sampling locations, to realize the data consistency block within a model-based deep learning image reconstruction scheme. This approach facilitates the joint and continuous optimization of the sampling pattern and the CNN parameters to improve image quality. We observe that the joint optimization of the sampling patterns and the reconstruction module significantly improves the performance of most deep learning reconstruction algorithms. The source code of the proposed joint learning framework is available at https://github.com/hkaggarwal/J-MoDL .

Deep learning

Image quality

Magnetic resonance imaging

parallel MRI

sampling

Reconstruction algorithms

Experiment design

Convolutional neural networks

Image reconstruction

Optimization

Compressed sensing

Details

Title: Subtitle: J-MoDL: Joint Model-Based Deep Learning for Optimized Sampling and Reconstruction
Creators: Hemant Kumar Aggarwal - Department of Electrical and Computer Engineering, University of Iowa, Iowa, IA, USA
Mathews Jacob - Department of Electrical and Computer Engineering, University of Iowa, Iowa, IA, USA
Resource Type: Journal article
Publication Details: IEEE journal of selected topics in signal processing, Vol.14(6), pp.1151-1162
DOI: 10.1109/JSTSP.2020.3004094
PMID: 33613806
PMCID: PMC7893809
NLM abbreviation: IEEE J Sel Top Signal Process
ISSN: 1932-4553
eISSN: 1941-0484
Publisher: IEEE
Grant note: 1R01EB019961-01A1; 1S10OD025025-01 / MRI Instrument
Language: English
Date published: 10/2020
Academic Unit: Roy J. Carver Department of Biomedical Engineering; Radiology; Electrical and Computer Engineering; Iowa Neuroscience Institute; Radiation Oncology
Record Identifier: 9984070374602771

Metrics

41 Record Views

79 Times Cited - Web of Science