English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Speed-of-sound (SoS) is an emerging ultrasound contrast modality, wherepulse-echo techniques using conventional transducers offer multiple benefits.For estimating tissue SoS distributions, spatial domain reconstruction fromrelative speckle shifts between different beamforming sequences is a promisingapproach. This operates based on a forward model that relates the sought localvalues of SoS to observed speckle shifts, for which the associated imagereconstruction inverse problem is solved. The reconstruction accuracy thushighly depends on the hand-crafted forward imaging model. In this work, wepropose to learn the SoS imaging model based on data. We introduce aconvolutional formulation of the pulse-echo SoS imaging problem such that theentire field-of-view requires a single unified kernel, the learning of which isthen tractable and robust. We present least-squares estimation of suchconvolutional kernel, which can further be constrained and regularized fornumerical stability. In experiments, we show that a forward model learned fromk-Wave simulations improves the median contrast of SoS reconstructions by 63%,compared to a conventional hand-crafted line-based wave-path model. Thissimulation-learned model generalizes successfully to acquired phantom data,nearly doubling the SoS contrast compared to the conventional hand-craftedalternative. We demonstrate equipment-specific and small-data regimefeasibility by learning a forward model from a single phantom image, where ourlearned model quadruples the SoS contrast compared to the conventionalhand-crafted model. On in-vivo data, the simulation- and phantom-learned modelsrespectively exhibit impressive 7 and 10 folds contrast improvements over theconventional model.

Learning the Imaging Model of Speed-of-Sound Reconstruction via a Convolutional Formulation