arXiv subjects

Samuel Beuret

Publications and source records attributed to Samuel Beuret.

3 recordsLinked to original sources

arXiv2026

Lens-Aware Differentiable Beamforming for In Vivo Distributed Aberration Correction with Curvilinear Transducers

This work extends ultrasound autofocusing via common midpoint phase error optimization to support curvilinear array geometries. Iterative model-based aberration correction via local sound speed estimation is performed by accounting for refraction caused by the transducer lens by using a differentiable bent-ray tracing approach. Model validation is performed in silico, using calibrated sound speed phantoms, and on in vivo human liver images. This work represents the first large-scale in vivo validation of our distributed aberration-correction method on 321 acquisitions from 81 high-BMI human liver subjects. In acquired images with anechoic regions, average contrast and CNR improved by $1.42 \pm 1.63$ dB (+18.4%), and $0.09 \pm 0.14$ (+10.3%), respectively. An average improvement was also observed across multiple image quality metrics: speckle brightness (+20.1%), coherence factor (+13.1%), lag-one coherence (+2.6%), common-midpoint correlation coefficient (+0.7%), and common-midpoint phase error (-9.1%, lower is better). All metric improvements were statistically significant. Additionally, a significant qualitative improvement in target structure and visibility was observed. These results demonstrate the potential for future clinical distributed aberration correction techniques using ultrasound autofocusing.

physics.med-ph↗

arXiv2026

A Wavefield Correlation Approach to Improve Sound Speed Estimation in Ultrasound Autofocusing

In pulse-echo ultrasound, aberration often degrades image quality when beamforming does not account for wavefront distortions. To address this issue, local sound speed estimators have been developed in the past decade for distributed aberration correction. Recently, methods based on iterative optimization have improved sound speed accuracy with respect to earlier approaches. However, the accuracy of these newer methods is limited by media with reverberation clutter and by the straight-ray model of wave propagation. To address these challenges, we propose using wavefield correlation (WFC) beamforming when performing sound speed optimization. WFC, an ultrasound adaptation of reverse time migration, correlates simulated forward-propagated transmit wavefields and backwards-propagated receive wavefields in order to reconstruct images. This process more accurately models wave propagation in heterogeneous media and can decrease diffuse clutter due to its spatiotemporal matched filtering effect. We implement herein a WFC beamformer using an auto-differentiation software and estimate the sound speed map by optimizing a regularized common-midpoint phase focusing criterion using gradient descent. This approach is compared to a previous method relying on delay and sum (DAS) with straight-ray time delay calculations on a variety of simulated, phantom, and in vivo data with large sound speed variations and clutter. Results show that using WFC decreases sound speed estimation error, leading to improvements in resolution and contrast in the corrected image. In particular, these promising results have potential to improve pulse-echo imaging for challenging clinical scenarios.

physics.med-ph↗

arXiv2018

Deep Learning of Human Perception in Audio Event Classification

In this paper, we introduce our recent studies on human perception in audio event classification by different deep learning models. In particular, the pre-trained model VGGish is used as feature extractor to process audio data, and DenseNet is trained by and used as feature extractor for our electroencephalography (EEG) data. The correlation between audio stimuli and EEG is learned in a shared space. In the experiments, we record brain activities (EEG signals) of several subjects while they are listening to music events of 8 audio categories selected from Google AudioSet, using a 16-channel EEG headset with active electrodes. Our experimental results demonstrate that i) audio event classification can be improved by exploiting the power of human perception, and ii) the correlation between audio stimuli and EEG can be learned to complement audio event understanding.

cs.SD↗