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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Ultrasound computed tomography (USCT) is actively being developed to quantifyacoustic tissue properties such as the speed-of-sound (SOS). Althoughfull-waveform inversion (FWI) is an effective method for accurate SOSreconstruction, it can be computationally challenging for large-scale problems.Deep learning-based image-to-image learned reconstruction (IILR) methods arebeing investigated as scalable and computationally efficient alternatives. Thisstudy investigates the impact of the chosen input modalities on IILR methodsfor high-resolution SOS reconstruction in USCT. The selected modalities aretraveltime tomography (TT) and reflection tomography (RT), which produce alow-resolution SOS map and a reflectivity map, respectively. These modalitieshave been chosen for their lower computational cost relative to FWI and theircapacity to provide complementary information: TT offers a direct -- while lowresolution -- SOS measure, while RT reveals tissue boundary information.Systematic analyses were facilitated by employing a stylized USCT imagingsystem with anatomically realistic numerical breast phantoms. Within thistestbed, a supervised convolutional neural network (CNN) was trained to mapdual-channel (TT and RT images) to a high-resolution SOS map. Moreover, the CNNwas fine-tuned using a weighted reconstruction loss that prioritized tumorregions to address tumor underrepresentation in the training dataset. Tounderstand the benefits of employing dual-channel inputs, single-input CNNswere trained separately using inputs from each modality alone (TT or RT). Themethods were assessed quantitatively using normalized root mean squared errorand structural similarity index measure for reconstruction accuracy andreceiver operating characteristic analysis to assess signal detection-basedperformance measures.

Investigating the Use of Traveltime and Reflection Tomography for Deep Learning-Based Sound-Speed Estimation in Ultrasound Computed Tomography