Passive acoustic mapping (PAM) is a key imaging technique for characterizing cavitation activity in therapeutic ultrasound applications. Recent model-based beamforming algorithms offer high reconstruction quality and strong physical interpretability. However, their computational burden and limited temporal resolution restrict their use in applications with time-evolving cavitation. To address these challenges, we introduce a PAM beamforming framework based on a novel convolutional formulation in the time domain, which enables efficient computation. In this framework, PAM is formulated as an inverse problem in which the forward operator maps spatiotemporal cavitation activity to recorded radio-frequency signals accounting for time-of-flight delays defined by the acquisition geometry. We then formulate a regularized inversion algorithm that incorporates prior knowledge on cavitation activity. Experimental results demonstrate that our framework outperforms classical beamforming methods, providing higher temporal resolution than frequency-domain techniques while substantially reducing computational burden compared with iterative time-domain formulations.

Sparse 2D arrays offer a cost-effective solution for 3D ultrasound imaging but suffer from limited image quality due to increased sidelobe and grating lobe artifacts, as well as reduced contrast. Traditional beamforming techniques are unable to consistently balance artifact suppression with speckle preservation. In this study we investigate the linear combination of different beamformer outputs to enhance sparse-array 3D imaging. A 3D encoder-decoder convolutional neural network was trained to predict a set of spatially varying weights for volumes generated by delay-and-sum (DAS), coherence-factor (CF) and a combination of filtered delay-multiply-and-sum (FDMAS) and CF. The model was trained exclusively on simulated phantoms with diverse echogenic structures and evaluated on a subset of simulated and experimental data. Results show that the proposed method achieves improved contrast and target detectability, while preserving speckle texture and producing smooth transitions between inclusions and surrounding tissue.

Ultrasound imaging faces a trade-off between image quality and hardware complexity caused by dense transducers. Sparse arrays are one popular solution to mitigate this challenge. This work proposes an end-to-end optimization framework that jointly learns sparse array configuration and image reconstruction. The framework integrates a differentiable Image Formation Model with a HARD Straight Thought Estimator (STE) selection mask, unrolled Iterative Soft-Thresholding Algorithm (ISTA) deconvolution, and a residual Convolutional Neural Network (CNN). The objective combines physical consistency (Point Spread Function (PSF) and convolutional formation model) with structural fidelity (contrast, Side-Lobe-Ratio (SLR), entropy, and row diversity). Simulations using a 3.5 MHz probe show that the learned configuration preserves axial and lateral resolution with half of the active elements. This physics-guided, data-driven approach enables compact, cost-efficient ultrasound probe design without sacrificing image quality, and it is expandable to 3-D volumetric imaging.

High-intensity focused ultrasound (HIFU) can induce cavitation, which requires monitoring for applications such as sonoporation, targeted drug delivery, and histotripsy. Passive acoustic mapping has been proposed as a cavitation monitoring method, but it suffers from limited axial resolution and coherent artifacts, particularly in non-inertial cavitation for the latter, leading to potential mislocalization of the cavitation cloud. Previous work introduced the Cross-Spectral Matrix Fitting (CMF), a regularized inverse problem approach, to improve axial resolution. Since coherence artifact positions vary with frequency, a common strategy is to mitigate them by summing maps from different frequencies. In this study, rather than summing frequency-specific maps, we propose Weighted Frequency Compounding-Cross-Spectral Matrix Fitting (WFC-CMF), which integrates multiple frequencies directly into the inverse problem and applies regularization to the reconstructed solution. To enhance robustness against outliers, we incorporate a Lorentzian M-estimator. Compared to Delay-and-Sum, WFC-CMF improves the Dice score by up to 35 points and outperforms summed-frequency CMF (sCMF) up to 14 points.

High-intensity focused ultrasound can induce cavitation, enabling therapeutic applications, such as sonoporation, targeted drug delivery, or histotripsy, which require accurate monitoring. Passive acoustic mapping, known as passive delay-and-sum (DAS), is a common cavitation monitoring technique but suffers from limited axial resolution. Recently, a regularized inverse approach, cross-spectral matrix fitting (CMF), incorporating prior knowledge into the reconstruction process, such as the group sparse characteristics of cavitation power maps (CMF-spTV), was proposed. While this approach improves axial resolution, it may fail to reconstruct cavitation clouds under specific conditions, e.g., elongated in the lateral direction. Herein, we propose a plug-andplay framework using a deep denoiser learned within a deep equilibrium framework. The proposed method improves the F1-score by 2 percentage points compared to CMF-spTV and by 32 points compared to DAS, with visibly better reconstructions, especially of laterally elongated cavitation clouds.

Realistic ultrasound simulation is essential for training and algorithm development. However, most existing methods generate single-frame images from thin slices of the probe's field of view, neglecting tissue deformation and spatial continuity across frames. We propose a physics-based simulation pipeline, fully compatible with any scatterer-based ultrasound simulator, which employs finite element modeling to update local scatterer maps under large-scale stability constraints, given the probe trajectory and tissue geometry. The framework produces continuous and anatomically accurate ultrasound sequences with physically consistent speckle evolution, providing a practical tool for training, visualization, and the development of freehand ultrasound reconstruction algorithms.

Passive acoustic mapping enables the spatial mapping and temporal monitoring of cavitation activity, playing a crucial role in therapeutic ultrasound applications. Most conventional beamforming methods, whether implemented in the time or frequency domains, suffer from limited axial resolution due to the absence of a reference emission onset time. While frequency-domain methods, the most efficient of which are based on the cross-spectral matrix, require long signals for accurate estimation, time-domain methods typically achieve lower spatial resolution. To address these limitations, we propose a linear model-based beamforming framework fully formulated in the time domain. The linear forward model relates a discretized spatiotemporal distribution of cavitation activity to the temporal signals recorded by a probe, explicitly accounting for time-of-flight delays dictated by the acquisition geometry. This model is then inverted using regularization techniques that exploit prior knowledge of cavitation activity in both spatial and temporal domains. Experimental results show that the proposed framework achieves enhanced or competitive cavitation map quality while using only 20\% of the data typically required by frequency-domain methods. This highlights the substantial gain in data efficiency and the flexibility of our spatiotemporal regularization to adapt to diverse passive cavitation scenarios, outperforming state-of-the-art techniques.

Despite being the widest used paradigm for echography, pulse-echo remains, even for ultrafast methods, an approach with intrinsic limitations. First, the framerate is constrained depending on the investigated depth. Second, it is blind to phenomena happening in the medium during most of the acquisition time. With the aim of imaging fast and short events, continuous emission ultrasound can be a solution as it overcomes latter issues by allowing to continuously probe the medium. However, the main challenge is then to extract wisely the information needed to estimate an echogenecity map of the medium at a given time. We propose a method to generate a B-Mode image sequence at a very high framerate, up to the sampling frequency of the system. To do so, the problem is reformulated as a sequence of synthetic aperture approaches. The carried out simulation study provides encouraging results with a tremendous temporal resolution enhancement.

Current imaging techniques in echography rely on the pulse-echo (PE) paradigm which provides a straight-forward access to the in-depth structure of tissues. They inherently face two major challenges: the limitation of the pulse repetition frequency, directly linked to the imaging framerate, and, due to the emission scheme, their blindness to the phenomena that happen in the medium during the majority of the acquisition time. To overcome these limitations, we propose a new paradigm for ultrasound imaging, denoted by continuous emission ultrasound imaging (CEUI) (Liebgott et al., 2023), for a single input single output (SISO) device. A continuous insonification of the medium is done by the probe using a coded waveform inspired from the radar and sonar literature. A framework coupling a sliding window approach (SWA) and pulse compression methods processes the recorded echoes to rebuild a motion-mode (M-mode) image from the medium with a high temporal resolution compared to state-of-the-art ultrafast imaging methods. A study on realistic simulated data, with regards to the motion of the medium, has been carried out and, achieved results assess an unequivocal improvement of the slow time frequency up to, at least, two orders of magnitude compared to ultrafast US imaging methods. This enhancement leads, therefore, to a ten times improvement in the temporal separability of the imaging system. In addition, it demonstrates the capability of CEUI to catch relatively short and quick events, in comparison to the imaging period of PE methods, at any instant of the acquisition.

Estimating myocardial fiber orientation is essential for understanding the heart's microstructure and diagnosing conditions like arrhythmias or heart failure. While Diffusion Tensor Imaging (DTI) is the current gold standard, it is often unavailable in many clinical settings due to cost and accessibility. Ultrasound offers a promising alternative, especially through methods that analyze the spatial coherence of RF signals, such as Backscatter Tensor Imaging (BTI). However, most existing techniques rely on correlation, which can miss important directional dependencies in the signals. In this work, we investigate the use of multivariate Granger causality (GC) as a complementary tool for estimating ultrasound-based fiber orientation. We first develop a simulation framework based on vector autoregressive models that allows independent control of correlation and causality. This framework enables systematic evaluation of correlation, time-domain GC, and frequency-domain GC under different signal regimes, including the effect of preprocessing such as bandpass filtering. We show that correlation performs well in correlation-dominant conditions, and GC is more effective when causal dependencies dominate. Frequency-domain GC in particular demonstrated robustness to filtering, whereas time-domain GC was more sensitive to preprocessing. The methodology was further validated experimentally using an in vitro nylon-fiber phantom, imaged with a 1024-element matrix array probe and a Verasonics Vantage 256 system. Coherent plane-wave compounding with 25 tilted transmissions was applied, and fiber orientations were varied from -60∘ to 60° using a rotation stage. Analysis of the compounded RF data confirmed the simulation findings: correlation provided stable estimates in regions dominated by instantaneous dependencies, while GC better captured directional patterns and was more sensitive to local structure. Balanced cases, observed both in simulation and experiment, suggested that real myocardial signals may arise from a mixture of correlated and causal interactions. Finally, fusing correlation and GC outputs through consensus and averaging improved robustness and accuracy of orientation estimation, especially in noisy or heterogeneous regions. While simulations relied on linear Gaussian VAR models, the phantom was more homogeneous than real tissue, and only in-plane orientations were studied, the results provide a foundation for future extensions toward 3D analysis, ex vivo validation, and real-time implementation. Overall, this study establishes multivariate Granger causality as a valuable complement to correlation for ultrasound-based fiber orientation estimation, demonstrating feasibility and promise as a cost-effective alternative to DTI with potential impact in cardiac diagnosis and monitoring.

Quantitative Acoustic Microscopy (QAM) is an imaging technology utilising high frequency ultrasound to produce quantitative two-dimensional (2D) maps of acoustical and mechanical properties of biological tissue at microscopy scale. Increased frequency QAM allows for finer resolution at the expense of increased acquisition times and data storage cost. Compressive sampling (CS) methods have been employed to produce QAM images from a reduced sample set, with recent state of the art utilising Approximate Message Passing (AMP) methods. In this paper we investigate the use of AMP-Net, a deep unfolded model for AMP, for the CS reconstruction of QAM parametric maps. Results indicate that AMP-Net can offer superior reconstruction performance even in its stock configuration trained on natural imagery (up to 63% in terms of PSNR), while avoiding the emergence of sampling pattern related artefacts.

This paper investigates statistical methods for analyzing the spatial coherence function to estimate myocardial fiber orientation, a critical factor in understanding cardiac microstructure and advancing cardiac imaging diagnostics. Four approaches, namely Correlation, Mutual Information, Kraskov Mutual Information, and Granger Causality, were evaluated using an in vitro experimental dataset designed to simulate myocardial fiber alignment. Our findings reveal that Granger Causality effectively captures complex and highly anisotropic structures, such as corners, while mutual information-based methods demonstrate superior stability and consistency across simpler regions. By applying a filter based on fractional anisotropy, the performance of each method was refined, highlighting their distinct strengths in fiber tracking. To leverage these complementary advantages, we propose the Fused and Consensus methods, which integrate the strengths of individual approaches to enhance coherence analysis and improve fiber orientation estimation. This study shows that the four methods complement each other by excelling in different aspects of fiber tracking, offering a robust framework for accurately characterizing fiber orientation. These insights can potentially improve the assessment of myocardial microstructure and aid in the early diagnosis and treatment of cardiac diseases.

Ultrasound attenuation imaging is gaining traction for its promising clinical diagnostic applications. Estimation methods such as Spatially Weighted Fidelity and Regularization Terms (SWIFT) and its deep learning-aided variant (DL-SWIFT) have demonstrated improved contrast-to-noise ratio (CNR) and more consistent attenuation coefficient slope (ACS) estimates through the use of spatially weighted formulations. However, both methods may still produce artifacts in heterogeneous regions with abrupt backscatter coefficient changes. To address these limitations, we propose ACS-Net, a deep unfolded Alternating Direction Method of Multipliers framework that integrates learned denoising operations within the classical iterative optimization process to reduce ACS estimation bias while preserving inclusion delineation. Phantom experiments confirmed a bias reduction of more than 40% on inclusions, while in vivo thyroid nodule results showed that ACS-Net decreases background coefficient of variation by nearly 50% and improves CNR by more than 30% compared to SWIFT and DL-SWIFT. These findings highlight the clinical promise of deep unfolded optimization methods for reliable and accurate ultrasound attenuation imaging.

Conventional ultrasound imaging relies on the pulse-echo (PE) paradigm, which inherently limits the frame rate and leaves the system blind during the receive period. To address these limitations, Continuous Emission Ultrasound Imaging (CEUI) has been proposed as a new approach that continuously insonifies the medium using a coded excitation. Owing to its continuous interaction with the medium, CEUI can in theory achieve a frame rate limited only by the sampling frequency. This work presents the first experimental validation of CEUI in a 1D setup using separate transmit and receive transducers for Motion-Mode imaging. Compared to PE performed at different PRFs (up to 10 kHz), CEUI achieved a frame rate of 1 MHz, enabling precise tracking of both slow and fast motions. These results demonstrate the feasibility of CEUI and its potential to surpass PE imaging in high frame rate applications.

Conventionally, ultrasound imaging techniques rely on pulse-echo (PE) acquisitions. PE paradigm provides a straight-forward access to the in-depth tissue structure, though limited for ultrafast phenomenon imaging. The first limitation is the pulse repetition frequency, directly proportional to the imaging framerate. Secondly, because of the small duty cycle, every region of the imaged medium does not produce any echo during most of the acquisition time. Continuous emission ultrasound imaging (CEUI) has been recently investigated to tackle these constraints for M-Mode and B-Mode imaging, relying on an uninterrupted insonification of the medium.Resulting recorded signals become, compared to PE, a "chaotic" spatio-temporal mixture of contribution from different regions in the field-of-view, backscattered at different times, and potentially received at the same time. Until now, to retrieve an image, proposed methods relied on a spatio-temporal dependent pulse compression for each estimated pixel, by looking for the signature of a short segment of the emitted waveform, in the received echoes. Repeating this process for a set of emitted segments enables to obtain images with a temporal resolution up to three orders of magnitude higher compared to ultrafast US imaging strategies with PE.This paper proposes a novel way to retrieve M-Mode images with continuous emission in order to tackle the trade-off between the gain in temporal resolution with its significant contrast degradation. An inverse problem approach enables to improve image consistency by estimating jointly all pixels from a same column of the M-Mode and enforcing temporal coherence between successive estimated states of the medium.

This paper introduces a beamforming technique for room acoustic analysis, contrasting from most conventional methods by operating in the time domain rather than the frequency domain. The technique estimates the direction of arrival of early reflections and the reverberant field using impulse responses recorded by a planar microphone array. A significant contribution is the design of a relaxed linear forward operator that models sound propagation based on the geometric setup. The proposal novelly leverages the impulsive sound waves’ sparsity in space and time simultaneously under an optimization framework. Comparative evaluations with conventional beamformers using synthetic and real data from a reverberant room validate the efficacy of the proposed technique for room acoustics analysis, highlighting its ability to discern early reflections temporally and estimate the reverberant field.

Quantitative acoustic microscopy (QAM) is a cutting-edge imaging modality that leverages very high-frequency ultrasound to characterize the acoustic and mechanical properties of biological tissues at microscopic resolutions. Radio-frequency signals are digitized and processed to yield two-dimensional maps. This paper introduces a weighted Hankel-based spectral method with a reweighting strategy to enhance robustness with regard to noise and reduce unreliable acoustic parameter estimates. Additionally, we derive, for the first time in QAM, Cramér-Rao bounds to establish theoretical performance benchmarks for acoustic parameter estimation. Simulations and experimental results demonstrate that the proposed method consistently outperform standard autoregressive approach under challenging conditions. These advancements promise to improve the accuracy and reliability of tissue characterization, enhancing the potential of QAM for biomedical applications.

The ultrasound backscatter coefficient (BSC) provides insight into the properties of acoustic scatterers for the quantitative assessment of tissue microstructure. The structure factor model (SFM) is an advanced model, specifically adapted for concentrated media, that allows the extraction of key scatterer parameters from BSC measurements. These include the radius a, volume fraction phi, and relative impedance contrast gamma(Z) which were found close to the values of cellular structures at high frequency (similar to 40 MHz). In this study, we focused on deriving the Cramer-Rao bound and comparing it with the mean squared error of the BSC parameters estimated using the SFM to understand the fundamental performance limits of the estimation process. We evaluated the impact of different parameter values on estimation performance in the presence of a Gaussian noise. The results indicate that estimation becomes more challenging for low values of gamma(Z), a, and phi.

Diffusion models for image generation have been a subject of increasing interest due to their ability to generate diverse, high-quality images. Image generation has immense potential in medical imaging because open-source medical images are difficult to obtain compared to natural images, especially for rare conditions. The generated images can be used later to train classification and segmentation models. In this paper, we propose simulating realistic ultrasound (US) images by successive fine-tuning of large diffusion models on different publicly available databases. To do so, we fine-tuned Stable Diffusion,(1) a state-of-the-art latent diffusion model, on BUSI (Breast US Images)(2) an ultrasound breast image dataset. We successfully generated high-quality US images of the breast using simple prompts that specify the organ and pathology, which appeared realistic to three experienced US scientists and a US radiologist. Additionally, we provided user control by conditioning the model with segmentations through ControlNet.(3) We will release the source code at http://code.sonography.ai/ to allow fast US image generation to the scientific community.

Ultrasound plane wave imaging is a cutting-edge technique that enables high frame-rate imaging. However, one challenge associated with high frame-rate ultrasound imaging is the high noise associated with them, hindering their wider adoption. Therefore, the development of a denoising method becomes imperative to augment the quality of plane wave images. Drawing inspiration from Denoising Diffusion Probabilistic Models (DDPMs), our proposed solution aims to enhance plane wave image quality. Specifically, the method considers the distinction between low-angle and high-angle compounding plane waves as noise and effectively eliminates it by adapting a DDPM to beamformed radiofrequency (RF) data. The method underwent training using only 400 simulated images. In addition, our approach employs natural image segmentation masks as intensity maps for the generated images, resulting in accurate denoising for various anatomy shapes. The proposed method was assessed across simulation, phantom, and in vivo images. The results of the evaluations indicate that our approach not only enhances image quality on simulated data but also demonstrates effectiveness on phantom and in vivo data in terms of image quality. Comparative analysis with other methods underscores the superiority of our proposed method across various evaluation metrics. The source code and trained model will be released along with the dataset at: http://code.sonography.ai.