Keywords: Cardiac Imaging, MRI Acceleration, End-to-end Deep Learning Pipeline, MRI Image Formation and Analysis
Abstract: Assessing cardiac health by measuring the cardiac function, for example using volume and ejection fractions, in cine magnetic resonance imaging is an essential step to assess the severity of cardiovascular diseases. However, motion artifacts caused by the difficulties the patients may have in either breath-holding or remaining still during acquisition, make the estimation of the segmentations required to compute the metrics above difficult, which in turn will undermine the quality of the estimated metrics. In this paper, we propose an end-to-end deep learning model that is optimized for two different tasks: reconstruction and segmentation. This is achieved by implementing a joint model that can achieve high segmentation accuracy while leveraging a fast acquisition by acting on under-sampled k-space data, under the assumption that some random motion occurs during cine cardiac MRI acquisition. Moreover, our joint model is able to reconstruct high quality images coupled with motion correction.
Registration: I acknowledge that acceptance of this work at MIDL requires at least one of the authors to register and present the work during the conference.
Authorship: I confirm that I am the author of this work and that it has not been submitted to another publication before.
Paper Type: novel methodological ideas without extensive validation
Primary Subject Area: Segmentation
Secondary Subject Area: Image Acquisition and Reconstruction
Confidentiality And Author Instructions: I read the call for papers and author instructions. I acknowledge that exceeding the page limit and/or altering the latex template can result in desk rejection.