Logo image
Dynamic Imaging Using Deep Bilinear Unsupervised Learning (Deblur)
Conference proceeding

Dynamic Imaging Using Deep Bilinear Unsupervised Learning (Deblur)

Abdul Haseeb Ahmed, Prashant Nagpal, Stanley Kruger and Mathews Jacob
2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI), pp.1099-1102
04/13/2021
DOI: 10.1109/ISBI48211.2021.9433882
PMCID: PMC8530343
PMID: 34691363
url
https://www.ncbi.nlm.nih.gov/pmc/articles/8530343View
Open Access

Abstract

Bilinear models such as low-rank and compressed sensing, which decompose the dynamic data to spatial and temporal factors, are powerful and memory efficient tools for the recovery of dynamic MRI data. These methods rely on sparsity and energy compaction priors on the factors to regularize the recovery. Motivated by deep image prior, we introduce a novel bilinear model, whose factors are regularized using convolutional neural networks. To reduce the run time, we initialize the CNN parameters by pre-training them on pre-acquired data with longer acquistion time. Since fully sampled data is not available, pretraining is performed on undersampled data in an unsupervised fashion. We use sparsity regularization of the network parameters to minimize the over-fitting of the network to measurement noise. Our experiments on free-breathing and ungated cardiac CINE data acquired using a navigated golden-angle gradient-echo radial sequence show the ability of our method to provide reduced spatial blurring as compared to low-rank and SToRM reconstructions.
Navigation Magnetic Resonance Imaging Storms Memory management Tools Reconstruction algorithms Spatial databases Cardiac MRI image reconstruction bilinear model dynamic imaging unsupervised learning

Details

Metrics

Logo image