No abstract

No abstract

Deep learning for ultrasound image formation is rapidly garnering research support and attention, quickly rising as the latest frontier in ultrasound image formation, with much promise to balance both image quality and display speed. Despite this promise, one challenge with identifying optimal solutions is the absence of unified evaluation methods and datasets that are not specific to a single research group. This paper introduces the largest known international database of ultrasound channel data and describes associated evaluation methods that were initially developed for the Challenge on Ultrasound Beamforming with Deep Learning (CUBDL), which was offered as a component of the 2020 IEEE International Ultrasonics Symposium. We summarize the challenge results and present qualitative and quantitative assessments using both the initially closed CUBDL evaluation test dataset (which was crowd-sourced from multiple groups around the world) and additional in vivo breast ultrasound data contributed after the challenge was completed. As an example quantitative assessment, single plane wave images from the CUBDL Task 1 dataset produced a mean generalized contrast-to-noise ratio (gCNR) of 0.67 and a mean lateral resolution of 0.42 mm when formed with delay-andsum beamforming, compared to a mean gCNR as high as 0.81 and a mean lateral resolution as low as 0.32 mm *CUBDL Organizers **Post-CUBDL Evaluator and Post-CUBDL Data Curator M.A.L.B. and A.W. acknowledge the National Institutes of Health (NIH) Trailblazer Award R21 EB025621 for partial support of their time on this project. H.R. and S.G. acknowledge the support of the Natural Sciences and Engineering Research Council of Canada (NSERC) RGPIN-2020-04612.

Highlights ASL is an alternative to 15 O-water for identifying hypoperfusion in FTD patients. ROI-based perfusion by ASL and 15 O-water were strongly correlated (R > 0.75). Hypoperfusion patterns identified by 15 O-water and ASL were in good agreement. Careful selection of the reference region is required to avoid erroneous results.

We present a method to efficiently separate signal in magnetic resonance imaging (MRI) into a base signal S0, representing the mainly T1-weighted component without T2*-relaxation, and its T2*-weighted counterpart by the rapid acquisition of multiple contrasts for advanced pharmacokinetic modelling. This is achieved by incorporating simultaneous multislice (SMS) imaging into a multi-contrast, segmented echo planar imaging (EPI) sequence to allow extended spatial coverage, which covers larger body regions without time penalty. Simultaneous acquisition of four slices was combined with segmented EPI for fast imaging with three gradient echo times in a preclinical perfusion study. Six female domestic pigs, German-landrace or hybrid-form, were scanned for 11 minutes respectively during administration of gadolinium-based contrast agent. Influences of reconstruction methods and training data were investigated. The separation into T1- and T2*-dependent signal contributions was achieved by fitting a standard analytical model to the acquired multi-echo data. The application of SMS yielded sufficient temporal resolution for the detection of the arterial input function in major vessels, while anatomical coverage allowed perfusion analysis of muscle tissue. The separation of the MR signal into T1- and T2*-dependent components allowed the correction of susceptibility related changes. We demonstrate a novel sequence for dynamic contrast-enhanced MRI that meets the requirements of temporal resolution (Δt < 1.5 s) and image quality. The incorporation of SMS into multi-contrast, segmented EPI can overcome existing limitations of dynamic contrast enhancement and dynamic susceptibility contrast methods, when applied separately. The new approach allows both techniques to be combined in a single acquisition with a large spatial coverage.