Dynamic imaging using a deep generative SToRM (Gen-SToRM) model

Qing Zou; Abdul Haseeb Ahmed; Prashant Nagpal; Stanley Kruger; Mathews Jacob

doi:10.1109/TMI.2021.3065948

Back

Dynamic imaging using a deep generative SToRM (Gen-SToRM) model

Journal article

Open access

Peer reviewed

Dynamic imaging using a deep generative SToRM (Gen-SToRM) model

Qing Zou, Abdul Haseeb Ahmed, Prashant Nagpal, Stanley Kruger and Mathews Jacob

IEEE transactions on medical imaging, Vol.40(11), pp.3102-3112

03/12/2021

DOI: 10.1109/TMI.2021.3065948

PMCID: PMC8590205

PMID: 33720831

Files and links (1)

url

https://www.ncbi.nlm.nih.gov/pmc/articles/8590205View

Open Access

Abstract

We introduce a generative smoothness regularization on manifolds (SToRM) model for the recovery of dynamic image data from highly undersampled measurements. The model assumes that the images in the dataset are non-linear mappings of low-dimensional latent vectors. We use the deep convolutional neural network (CNN) to represent the non-linear transformation. The parameters of the generator as well as the low-dimensional latent vectors are jointly estimated only from the undersampled measurements. This approach is different from traditional CNN approaches that require extensive fully sampled training data. We penalize the norm of the gradients of the non-linear mapping to constrain the manifold to be smooth, while temporal gradients of the latent vectors are penalized to obtain a smoothly varying time-series. The proposed scheme brings in the spatial regularization provided by the convolutional network. The main benefit of the proposed scheme is the improvement in image quality and the orders-of-magnitude reduction in memory demand compared to traditional manifold models. To minimize the computational complexity of the algorithm, we introduce an efficient progressive training-in-time approach and an approximate cost function. These approaches speed up the image reconstructions and offers better reconstruction performance.

Magnetic Resonance Imaging

Time Series Analysis

Manifolds

Deep image prior

CNN

Storms

Imaging

Generators

Manifold approach

Unsupervised learning

Electronics packaging

Generative model

Details

Title: Subtitle: Dynamic imaging using a deep generative SToRM (Gen-SToRM) model
Creators: Qing Zou - Applied Mathematics and Computational Sciences (AMCS) program at the University of Iowa, Iowa City, USA. (e-mail: zou-qing@uiowa.edu)
Abdul Haseeb Ahmed - Department of Electrical and Computer Engineering, University of Iowa, Iowa City, USA
Prashant Nagpal - Department of Radiology, University of Iowa, Iowa City, USA
Stanley Kruger - Department of Radiology, University of Iowa, Iowa City, USA
Mathews Jacob - Department of Electrical and Computer Engineering, University of Iowa, Iowa City, USA
Resource Type: Journal article
Publication Details: IEEE transactions on medical imaging, Vol.40(11), pp.3102-3112
DOI: 10.1109/TMI.2021.3065948
PMID: 33720831
PMCID: PMC8590205
NLM abbreviation: IEEE Trans Med Imaging
ISSN: 0278-0062
eISSN: 1558-254X
Publisher: Institute of Electrical and Electronics Engineers
Grant note: 1S10OD025025-01; R01EB019961 / National Institutes of Health (10.13039/100000002) 01AG067078-01A1 / National Institute on Aging (10.13039/100000049)
Language: English
Electronic publication date: 03/12/2021
Date published: 03/12/2021
Academic Unit: Roy J. Carver Department of Biomedical Engineering; Radiology; Electrical and Computer Engineering; Iowa Technology Institute; Iowa Neuroscience Institute; Radiation Oncology
Record Identifier: 9984070014502771

Metrics

24 Record Views

33 Times Cited - Web of Science