5 - A Deep-Neural-Network-based Signal Invariance Approach to overcome Unresolved Cardiac MR artifacts exhibited in Training Data

Friday, February 8

11:22 AM - 11:25 AM

Location: Grand C

Presenter(s)

YS

Yudai Suzuki

Graduate Student
Gifu University

Author Only(s)

NB

Nanyque Boyd

Combined Degree Student
Illinois Institute of Technology
MV

Marcella Vaicik, PhD

Research Assistant Professor
Illinois Institute of Technology
Amit Patel, MD

Associate Professor of Medicine and Radiology
University of Chicago
SH

Satoru Hayamizu, PhD

Professor
Gifu University
ST

Satoshi Tamura, PhD

Associate Professor
Gifu University
Keigo Kawaji, PhD

Assistant Professor
Illinois Institute of Technology

Background:

Novel MRI approaches using Deep Neural Networks^1-3 (DNN) typically assume: 1) high-fidelity of training data (i.e. fully sampled, artifact-free images), and 2) optimized DNN models. For CMR, clinical imaging constraints, including subject motion and systems hardware imperfections may yield suboptimal training data. We propose a new reconstruction approach using DNN that: a) accounts for imperfect CMR training data fidelity, and b) does not require a carefully tuned DNN model. Our approach deliberately exploits training-derived remnant artifacts as vectorized projections that are separable from signal source. Using non-Cartesian high-temporal Cine-CMR^4,5 (~15ms/phase) example, we demonstrate one application of our generalizable framework that may mitigate conditions for training data fidelity and need for elaborate DNN modeling.

Methods:

In brief, our generalized approach incorporates DNN-based training that deliberately projects unresolved reconstruction artifacts onto known artifact-vectors under a mathematical transform domain; we thereby separate orthogonally aligned artifact-vectors from the invariant signal source. For this Radial cine-CMR example (n=10 subjects; 14 slices x 80 cardiac phases; 9-training+1-validation cycled for every subject), we employed for DNN training our Nyquist-satisfied reconstructions that employed a weighted [KWIC] temporal window⁶ that nonetheless contained unresolved reconstruction artifacts. R=5 accelerated re-gridded data was used for DNN input. Schematic of artifact vector projection example is shown in Figure 1, derived via DNN in the image x- and y-directions for illustrative clarity. These artifact vectors can be yielded in two orthogonal directions, and filtered from the invariant source (e.g. LPF, blind-source separation [PCA, ICA], etc.). For the proposed approach assessment, invariance of Gaussian noise distribution was first examined in every output frame using the Kolmogorov-Smirnov (KS) test⁷. A per-cardiac phase apparent Signal-to-Noise-Ratio (aSNR) map was next generated to yield a fair map ratio for comparisons between alternative reconstructions⁴.

Results:

Figure 2a shows the KS-test map that satisfied the Gaussian noise condition for the subsequent SNR-map-based assessment in a fair and controlled manner across alternative reconstruction techniques. Figure 2b shows: a) input, b) training, c) previous⁴, and d) proposed methods. The derived aSNR-map ratios across 10x14x80 frames were: 10.33 (proposed/training); 5.33 (proposed/previous). Both comparisons were statistically significant (p<0.0005).

Conclusion:

We demonstrate approach that exploits signal invariance and artifact vectorization for CMR using imperfectly trained DNNs. This example yielded high-frame-rate cine-CMR reconstructions with superior radial streaking artifact suppression compared to previous methods, despite the presence of unresolved radial streaking in the employed training dataset as answer key.