Photoacoustics

Light-emitting diode (LED)-based photoacoustic (PA) imaging offers a compact, safe, and cost-effective alternative to laser systems. However, low LED power leads to low optical fluence, which in turn limits penetration depth. We report a systematic optimization of LED-PA performance by jointly tuning ultrasound (US) probe frequency and LED pulse width, validated in both phantom and in vivo studies. Using a commercially available LED-PA/US platform, we compared a custom 5 MHz transducer with commercial 7 and 10 MHz probes under LED pulses of 30 to 100 ns. The 5 MHz probe with a 100 ns pulse achieved the best trade-off between depth sensitivity and resolution, enabling detection of targets up to 18 mm. In vivo experiments demonstrated, for the first time, clear visualization of the carotid artery and deep-seated breast vessels using LED-based PA imaging. These findings show that careful optimization of probe frequency and pulse width can substantially extend the depth performance of LED-PA, advancing its potential for vascular and oncologic applications.

SignificanceConventional spectral photoacoustic imaging (sPAI) to assess tissue oxygenation (sO2) uses optical wavelengths in the first near infrared window (NIR-I). This limits the maximum imaging depth ([~]1 cm) due to high spectral coloring of biological tissues. AimSecond near infrared or short-wave infrared (NIR-II or SWIR) wavelengths (950-1400 nm) show potential for deep tissue sPAI due to the exponentially reduced tissue scattering and higher maximum exposure threshold (MPE) in this wavelength range. However, to date, a systematic assessment of NIR-II wavelengths for sPAI of tissue sO2 has yet to be performed. ApproachThe NIR-II PA spectra of oxygenated and deoxygenated hemoglobin was first characterized using a phantom. Optimal wavelengths to minimize spectral coloring were identified. The resulting NIR-II PA imaging methods were then validated in vivo by measuring renal sO2 in adult female rats. ResultssPAI of whole blood under a phantom and of circulating renal blood in vivo, demonstrated PA spectra proportional to wavelength-dependent optical absorption. NIR-II wavelengths had a [~]50% decrease in error of spectrally unmixed blood sO2 compared to conventional NIR-I wavelengths. In vivo measurements of renal sO2 validated these findings and demonstrated a [~]30% decrease in error of estimated renal sO2 when using NIR-II wavelengths for spectral unmixing in comparison to NIR-I wavelengths. ConclusionssPAI using NIR-II wavelengths improved the accuracy of tissue sO2 measurements. This is likely due to the overall reduced spectral coloring in this wavelength range. Combined with the increased safe skin exposure fluence limits in this wavelength range, demonstrate the potential to use NIR-II wavelengths for quantitative sPAI of sO2 from deep heterogeneous tissues.

Ultrasound localization microscopy (ULM) has revolutionized microvasculature imaging by surpassing the diffraction limit via microbubbles. ULM demonstrates exceptional potential to resolve micrometer-scale vascular structures in both preclinical and clinical studies. However, its performance evaluation remains challenging primarily due to the lack of reference microvascular phantoms featuring micrometer-scale, hierarchical branches, and realistic vascular structures. Inspired by microfluidic chip techniques, we present an organ-on-a-chip protocol for fabricating agarose-based micro-vessel network phantoms with ground truth. The vasculature pattern offers design versatility, enabling on-demand customization. We experimentally demonstrated the feasibility of the vasculature phantom using two adapted patterns. The first was a leaf pattern, which exhibited intrinsic quasi-two-dimensional venation network with hierarchical and branching channels similar to animal vasculature. The second was a kidney pattern, which was based on a two-dimensional projection of real human vasculature obtained from micro computed tomography. The microbubble solution was perfused into the phantoms by capillary force and gravity. The ULM-reconstructed vasculature maps agreed well with the ground truth. ULM achieved a high sensitivity of 0.97 and 0.95, but a low precision of 0.37 and 0.60, for the leaf and kidney phantom, respectively. The results indicated the capability of ULM to reconstruct vessel structures while making many false positive predictions. The proposed protocol holds significant promise for the development and optimization of ultrasound microvascular imaging techniques.

This study presents the use of sub-micron nanobubbles (NBs) as contrast agents for ultrasound localization microscopy (ULM), a super-resolution imaging technique that visualizes microvascular structure and flow beyond the acoustic diffraction limit. While ULM has traditionally relied on micron-sized microbubbles (MBs), the reduced dimensions and prolonged circulation times of NBs make them attractive candidates for localization-based imaging. However, their weaker acoustic responses present significant challenges for reliable detection and tracking. To address this challenge, we developed the ULM Master GUI, an interactive framework for optimization of the complete ULM processing pipeline. Using custom ultrasound-compatible wall-less gelatin flow phantoms containing vessel-mimicking channels and bifurcations ranging from 100 to 500 m, we demonstrate that NB-based ULM achieves velocity reconstruction and flow partitioning measurements comparable to conventional MB-based ULM. Across all investigated geometries, NBs faithfully reproduced the underlying flow patterns and hemodynamic behavior despite their substantially reduced acoustic scattering. These findings establish the feasibility of NB-based ULM, expand the range of contrast agents available for localization microscopy, and provide a foundation for future super-resolution ultrasound imaging using nanoscale acoustic contrast agents. The ULM processing GUI is publicly available at https://github.com/grisha1998/ulm-super-resolution-toolbox.

The fast evolution of medical micro- and nanorobots in the endeavor to perform non-invasive medical operations in living organisms boosted the use of diverse medical imaging techniques in the last years. Among those techniques, photoacoustic (PA) tomography has shown to be promising for the imaging of microrobots in deep-tissue (ex vivo and in vivo), as it possesses the molecular specificity of optical techniques and the penetration depth of ultrasound imaging. However, the precise maneuvering and function control of microrobots, in particular in living organisms, demand the combination of both anatomical and functional imaging methods. Therefore, herein, we report the use of a hybrid High-Frequency Ultrasound (HFUS) and PA imaging system for the real-time tracking of magnetically driven micromotors (single and swarms) in phantoms, ex vivo, and in vivo (in mice bladder and uterus), envisioning their application for targeted drug-delivery.

In vivo infrared thermography is limited by the inherently poor spatial resolution at long wavelengths, low contrast, and the lack of biocompatible contrast agents. Here, we present 3-5 m mid-wave infrared (MWIR) thermography enhanced by an artificial intelligence (AI) network and cold phosphate-buffered saline (PBS) as a thermal contrast agent for noninvasive in vivo imaging with high contrast and resolution. MWIR imaging enabled high thermal sensitivity with microscale spatial resolution, strong relative thermal contrast, and facilitated visualization of the subcutaneous vasculature in the human arm, hand, ankle, the femoral artery and vein in rats, and the femoral vessels in mice, with image contrast further enhanced by AI networks. In a 4T1 tumor-bearing mouse model, AI-enhanced MWIR resolved early-stage tumors of ~2.3 mm and metastases as small as ~1.7 mm. Using cold PBS as a MWIR thermal contrast agent, we achieved precise tumor boundary visualization and real-time imaging-guided tumor resection. AI-enhanced MWIR offers a promising solution for early diagnosis and improved surgical precision.

AbstractO_ST_ABSSignificanceC_ST_ABSPhotoacoustic (PA) imaging has garnered considerable attention due to its capability to render vascular images in a label-free manner. Specifically, devices employing a hemispherical detector array (HDA) have been heralded for various clinical applications, owing to their potential to yield high reproducibility three-dimensional images. While high-resolution models utilizing high-frequency sensors have been introduced for animal experimentation, their evaluation has been constrained to a single wavelength. In this study, we demonstrate the applicability of in vivo mouse models for visualizing body oxygen saturation distribution using dual wavelengths. AimWith the aid of our uniquely developed device and analysis software, our primary objective is to map the spatial distribution of the hemoglobin oxygen saturation coefficient (S-factor) through non-invasive in vivo imaging. Subsequently, we aim to observe the temporal alterations within this distribution, specifically assessing changes in hemoglobin oxygen saturation in both normal and tumor vessels over time. ApproachHigh-quality S-factor images were obtained by integrating a newly developed scanning sequence for high contrast with alternate two-wavelength irradiation. Following validation with phantoms, in vivo images were procured in mice. Sequential scanning of the same mouse yielded information about temporal changes. S-factor evaluation was conducted with our photoacoustic image viewer to analyze trends in hemoglobin oxygen saturation. ResultsHigh-contrast images were achieved by increasing the number of integrations during scanning. S-factor images were acquired using both healthy and tumor-bearing mice. Vessels within the liver and kidneys were distinctly reconstructed, and differences in oxygen saturation discriminated between arteries and veins. Repeated measurements on the same mice, both live and post-euthanasia, provided spatiotemporal information, such as a decrease in oxygen saturation after euthanasia or a precipitous drop in oxygen saturation inside the tumor nine days post-cell line transplantation. ConclusionsBy analyzing S-factor images using a photoacoustic imaging system designed for animal experiments, we succeeded in discerning variations in in vivo oxygen saturation. The custom-built system holds promise as a versatile tool for diverse basic research endeavors, as it can seamlessly interface with human clinical applications.

Lactate is a key metabolic driver in oncology and immunology. Even in the presence of physiological oxygen levels, most cancer cells upregulate anaerobic glycolysis, resulting in abnormal lactate production and accumulation in the tumor microenvironment. The development of more effective, sensitive, and safe probes for detecting extracellular lactate holds the potential to impact cancer metabolic profiling and staging significantly. Macrocyclic-based PARACEST agents have been reported to act as shift reagents (SRs) and detect extracellular lactate via CEST MRI. Here, we introduce a new family of SRs based on the PCTA ligand, an inherently stable and kinetically inert group of molecules with the potential for (pre)clinical translation. We observed that Yb-PCTA and Eu-PCTA can significantly shift lactate -OH signals in CEST spectra. In vitro, CEST MRI experiments proved that imaging extracellular lactate with these complexes is feasible and maintains high specificity even in the presence of competing small metabolites in blood and the tumor microenvironment. In vivo, preclinical imaging demonstrated that Yb-PCTA can be safely administered intravenously in mice to detect extracellular lactate non-invasively. This work represents a significant step towards precision and molecular imaging, demonstrating that the PCTA-ligand is a promising scaffold for developing molecular imaging sensors. These chemical sensors have broad medical applications, particularly in oncology and physiology.

The resolution of 3D Ultrasound Localization Microscopy (ULM) is determined by acquisition parameters such as frequency and transducer geometry but also by microbubble (MB) concentration, which is also linked to the total acquisition time needed to sample the vascular tree at different scales. In this study, we introduce a novel 3D anatomically- and physiologically-realistic ULM simulation framework based on two-photon microscopy (2PM) and in-vivo MB perfusion dynamics. As a proof of concept, using metrics such as MB localization error, MB count and network filling, we could quantify the effect of MB concentration and PSF volume by varying probe transmit frequency (3-15 MHz). We find that while low frequencies can achieve sub-wavelength resolution as predicted by theory, they are also associated with prolonged acquisition times to map smaller vessels, thus limiting effective resolution. A linear relationship was found between maximal MB concentration and inverse point spread function (PSF) volume. Since inverse PSF volume roughly scales cubically with frequency, the reconstruction of the equivalent of 10 minutes at 15 MHz would require hours at 3 MHz. We expect that these findings can be leveraged to achieve effective reconstruction and serve as a guide for choosing optimal MB concentrations in ULM.

Non-invasive, high-resolution visualization of mouse brain vasculature remains challenging due to significant light scattering and absorption by mammalian tissues, hence many optical imaging protocols require scalp and/or skull excision. Here we present a fully reversible tartrazine-based optical clearing strategy that enables cortex-wide optical coherence tomography angiography (OCTA) through intact scalp and skull. We characterized tartrazine properties in the near infrared (NIR)-II band of the 1.3 {micro}m swept-source OCTA system, confirming minimal absorption across 1.25-1.35 {micro}m wavelength range and an effectively constant refractive index, suggesting negligible OCTA distortions. Spatially selective agent application showed that intracranial vessels emerge selectively within the treated region of interest (ROI), whereas untreated regions retain strong interference by the scalp vascular features. Depth-encoded projections and cross-sectional OCTA demonstrated an increased signal recovery at depth and an extended vessel-detection range after clearing. Vessel-map changes were quantified using intersection-over-union and Dice coefficients, yielding high similarity outside the ROI and reduced similarity within the ROI, consistent with a transition from scalp to brain vasculature. Reproducibility was confirmed in three independent 11-week-old mice and validated against scalp-removed reference OCTA. Screening tartrazine in the 0.3-0.8 Molar concentration range (7-min application) identified 0.6 M as optimal for whole-cortex scanning, balancing clearing efficacy and solution handling. Finally, the protocol generalized across mice aged 5-18 weeks. This approach provides a practical route to non-invasive structural cerebrovascular mapping with OCTA.

Objective: This study aims to develop and validate a quantitative, depth-resolved anisotropy imaging framework that extends ARFI-based focal degree-of-anisotropy (DoA) estimation into two-dimensional mapping by modeling the depth-dependent relationship between shear modulus ratio (SMR) and peak displacement ratio (PDR). Methods: We propose APRIL (Adaptive Polynomial Regression for anisotropy Imaging via ARFI-induced DispLacements), a framework for quantitative, depth-resolved DoA imaging that adaptively selects polynomial regression or shape-preserving spline interpolation based on excitation PSF asymmetry. Training data were generated using an LS-DYNA3D + Field II simulation pipeline in homogeneous transversely isotropic media (SMR 0.9-4.9). Testing included shifted SMRs under varied acoustic conditions and three heterogeneous inclusion configurations (anisotropic inclusion in isotropic background and vice versa). Experimental validation was performed in an in-vivo murine tumor model over the time, ex-vivo chicken breast, and tissue-mimicking gelatin phantoms, using a Verasonics system with an L11-5v transducer. Results: APRIL achieved depth-resolved SMR prediction errors below 9% over 10-30 mm, with highest accuracy in the focal region (MAE 2.3%, RMSE < 0.1) and stable performance across PSF transition zones. In heterogeneous phantoms, it reconstructed anisotropy maps with SSIM up to 86% and MPE below 7%, accurately delineating inclusion boundaries. Under acoustic parameter variations, mean absolute errors remained below 10%, demonstrating robustness to system and tissue heterogeneity. Conclusion: APRIL enables robust, two-dimensional anisotropy imaging beyond focal estimates. Significance: The method provides a physically grounded and generalizable framework for clinically viable anisotropy biomarkers in muscle, tendon, kidney, tumor and breast tissues.

Radiofrequency (RF) ablation is a minimally invasive therapy for heart arrhythmia, including atrial fibrillation (A-fib), which creates lesions using an electric current to isolate the heart from abnormal electrical signals. However, conventional RF procedures do not involve intraoperative monitoring of the area and extent of ablation-induced necrosis, making the assessment of the procedure completeness challenging. Previous studies have suggested that spectroscopic photoacoustic (sPA) imaging is capable of differentiating ablated tissue from its non-ablated counterpart based on PA spectrum variation. Here, we aim to demonstrate the applicability of sPA imaging in an in vivo environment, where the cardiac motion presents, and introduce a framework for mapping the necrotic lesion using cardiac-gated sPA imaging. We computed the degree of necrosis, or necrotic extent (NE), by dividing the quantified ablated tissue contrast by the total contrast from both ablated and non-ablated tissues, visualizing it as continuous colormap to highlight the necrotic area and extent. To compensate for tissue motion during the cardiac cycle, we applied the cardiac-gating on sPA data, based on the image similarity. The in vivo validation of the concept was conducted in a swine model. As a result, the ablation-induced necrotic lesion at the surface of the beating heart was successfully depicted throughout the cardiac cycle through cardiac-gated sPA (CG-sPA) imaging. The results suggest that the introduced CG-sPA imaging system has great potential to be incorporated into clinical workflow to guide ablation procedures intraoperatively.

Accurate signal detection of ultrasound contrast agents, such as microbubble (MB) and gas vesicle (GV), in the presence of clutter and noise is essential to increase image quality in super-resolution ultrasound imaging (SRUS) and achieve precise GV localization. We developed and evaluated an eigen-image based signal detection method using singular value decomposition (SVD) and changepoint detection to automatically segment the data that are closely related to physical events such as MB flow and GV collapse. Eigen-image based method was compared with the elbow point and hard thresholding method when selecting MB signals after SVD of raw data, acquired from phantom and in vivo experiments. Image reconstructed by eigen-image based method was also compared with unregistered difference image for GV localization when moving GVs in a phantom were collapsed by ultrafast plane waves. The eigen-image based MB signal detection method resulted in higher vessel density (VD) visualization in both the phantom and in vivo mouse tumor. It also achieved increased signal-to-noise ratio (SNR) in both cases. Moreover, this method localized moving GVs more efficiently than the difference imaging method, without requiring pixel registration based on landmarks. The eigenimage based method offers a reliable and automated approach to MB and GV signal detection for both SRUS and point target localization. This approach is a valuable tool for medical imaging providing high-quality vessel images along with accurate locations of moving ultrasound contrast agents, which can be potentially translatable to clinical diagnosis and pre-clinical research. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=104 SRC="FIGDIR/small/662077v1_ufig1.gif" ALT="Figure 1"> View larger version (30K): org.highwire.dtl.DTLVardef@d03b0corg.highwire.dtl.DTLVardef@d4b318org.highwire.dtl.DTLVardef@3a2894org.highwire.dtl.DTLVardef@3e2962_HPS_FORMAT_FIGEXP M_FIG C_FIG HighlightsO_LIAn approach to efficiently detect point targets including microbubbles and gas vesicles is developed using eigenimages that describe the changes of point targets over time for super-resolution ultrasound imaging and gas vesicle localization. C_LIO_LISignals from point targets were automatically identified from clutter and noise by the changepoint detection method without prior Information. C_LIO_LIThis approach enhances microvasculature visualization of tumors in mice and improves the localization of gas vesicles in a phantom. C_LI

Viscoelastic characterization of tissue has significant diagnostic value in oncology, as tumor progression alters both elasticity and viscosity in ways that neither property alone can fully capture. Existing acoustic radiation force (ARF)-based methods such as Viscoelastic Response (VisR) ultrasound estimate relative elasticity and viscosity through per-A-line nonlinear model fitting, which is computationally intensive and requires auxiliary simulations to correct elasticity-dependent bias. This work presents VESTA (Machine Learning-Enabled Estimation of ViscoElastic Ratios from On-Axis Spatio-Temporal ARFI Features), a two-stage data-driven pipeline that predicts elasticity ratio (ER) and viscosity ratio (VR) directly from seven normalized ARFI displacement features at the A-line level, without model fitting or compensation. Stage~1 is an MLP classifier that detects inclusion boundaries from normalized peak displacement and negative peak velocity ratios; Stage~2 is a dilated Conv1D regression model that estimates ER and VR along the full axial sequence using the predicted mask alongside displacement features. The pipeline was trained on 500 simulated inclusion scenarios spanning three geometries, five focal depths, two F-numbers, and a broad range of material contrasts. In silico, mean predicted ER and VR were within 12\% of ground truth across all geometries, with performance best when ER and VR were moderate or decoupled. Experimental validation on a chicken breast phantom demonstrated plausible generalization to real tissue heterogeneity. Applied to an in vivo murine 4T1 breast cancer model, the pipeline tracked treatment-related attenuation of mechanical contrast in paclitaxel-treated tumors relative to controls over a 36-day imaging period, supporting its relevance for tumor monitoring.

Understanding brain-wide hemodynamic responses to different stimuli at high spatiotemporal resolutions can help study neuro-disorders and brain functions. However, the existing brain imaging technologies have limited resolution, sensitivity, imaging depth and provide information about only one or two hemodynamic parameters. To address this, we propose a multimodal functional ultrasound and photoacoustic (fUSPA) imaging platform, which integrates ultrafast ultrasound and multispectral photoacoustic imaging methods in a compact head-mountable device, to quantitatively map cerebral blood volume (CBV), cerebral blood flow (CBF), oxygen saturation (SO2) dynamics as well as contrast agent enhanced brain imaging with high spatiotemporal resolutions. After systematic characterization, the fUSPA system was applied to quantitatively study the changes in brain hemodynamics and vascular reactivity at single vessel resolution in response to hypercapnia stimulation. Our results show an overall increase in brain-wide CBV, CBF, and SO2, but regional differences in singular cortical veins and arteries and a reproducible anti-correlation pattern between venous and cortical hemodynamics, demonstrating the capabilities of the fUSPA system for providing multiparametric cerebrovascular information at high-resolution and sensitivity, that can bring insights into the complex mechanisms of neurodiseases.

Skin colour is known to confound readouts from optical devices that make measurements through the skin, which can adversely impact the care of patients with darker skin. Photoacoustic imaging (PAI) is making its way from the laboratory to the clinic, however, combining optics and ultrasound for deep tissue imaging leads to a complex relationship between photoacoustic-derived imaging biomarkers and skin melanin concentration. Furthermore, no generalisable correction of the confounding effects of skin colour in PAI has been demonstrated. We sought to overcome this limitation by recruiting a healthy volunteer cohort with the most diverse range of skin tones ever assembled in the field, with participants from Fitzpatrick types I to VI and with vitiligo. From this comprehensive dataset, we identified and characterised two physical mechanisms responsible for skin colour-dependent degradation in both image quality and biomarker quantification. Accompanied by detailed theoretical modelling, we demonstrated that strong light absorption by melanin leads to spectral colouring, which dominates in individuals with low skin melanin pigmentation. We further identified the backscattering of ultrasound waves generated in the skin as a major source of image artefacts for individuals with high skin melanin pigmentation. With this improved understanding of the physical basis, we were able to develop a fast and practicable correction method for spectral colouring and adapted a plane-wave ultrasound reconstruction algorithm to reveal the ultrasound scatterer distribution encoded in the photoacoustic timeseries. Our findings highlight the need for more advanced image reconstruction methods to enable equitable clinical application of PAI. One Sentence SummaryPhotoacoustic imaging is proven to suffer from measurement inaccuracies in people with darker skin, which could adversely impact patient care if not appropriately corrected.

PurposeIn this work, we aimed to establish a flow phantom for multi-modal PET and spectral CT imaging to improve blood flow quantification. MethodsA modular flow phantom was built with materials compatible with both PET and spectral CT. A peristaltic pump was used to allow for recirculation. Pores were installed through the aorta to allow for tissue exchange between the blood and tissue compartments, and valves were placed in line with the aorta to control the pressure gradient between compartments. We characterized the system using saline bolus experiments, dynamic PET imaging, and iodine-based spectral CT acquisitions. A blood flow (K1) of 1.0 mL/min/mL with a pressure range of approximately 1.0-3.0 psi was targeted. Using compartmental modeling, we estimate K1 across phantom configurations and evaluate the consistency of perfusion-related parameters derived from saline, PET, and spectral CT measurements. ResultsWith the four pore, two valve configuration, target K1 of 1.0 mL/min/mL was achieved with a physiologic pressure range (2.2-3.5 psi) and a pump speed of 150 rpm. Further, the flow phantom was also able to recapitulate K1 across a range of values through adjustable modifications to the phantom configuration. ConclusionsWe present a modular multimodal flow phantom with a tissue-mimicking compartment, vascular tubing with an adjustable number of pores and valves, and 3D-printed components to support tunable exchange between blood-pool and tissue compartments and controlled dynamic perfusion imaging with same-session PET and spectral CT. Such a setup will enable the development of multi-modal approaches for evaluating tissue perfusion.

ObjectiveSuper-resolution ultrasound (SR-US) reveals microvascular structures with exquisite resolution, but clinical translation remains limited by the need for ultrafast frame rates, massive data volumes, and long reconstruction times. This work proposes a deep learning framework that reconstructs microvascular maps from low-frame-rate enhanced ultrasound sequences, bypassing explicit microbubble localization and tracking. MethodsA transformer-decoder network with learned linear projections was designed to model spatiotemporal dependencies across sparse contrast-enhanced ultrasound sequences and reconstruct vessel probability maps, refined via a post-processing enhancement stage. Single-head self-attention captures temporal correlations under challenging conditions including overlapping microbubbles and low signal-to-noise ratios. Binary cross-entropy loss guided training to preserve vascular topology across synthetic and in vivo datasets. In vivo rat brain bolus data from the PALA challenge was used to evaluate this approach under up to 500 - fold data reduction (341 frames at 2 FPS vs. 170400 frames at 1000 FPS in standard ULM). ResultsDespite aggressive undersampling, the proposed pipeline recovered coherent microvascular architecture where conventional ULM pipelines applied to the same sparse data failed to produce continuous vascular networks. Major branches and higher-order microvessels remained visible with apparent vessel widths broadened by approximately three-fold relative to reference SR-US. End-to-end reconstruction completed in 28-133 seconds on an NVIDIA H100 GPU depending on the number of frames employed. ConclusionThe reported approach preserved vascular topology with fast reconstruction and low data overhead, albeit at lower resolution. The substantial reduction in frames and computation time highlights the translational potential of this SR-US-inspired microvascular imaging approach.

Brain perfusion relies on a complex vascular network of arteries, veins, and capillaries to meet its constant demand for oxygen and nutrients. Disruption of this microvascular system is a hallmark of many neurological disorders, including small vessel disease, stroke, and brain tumors. As such, high-resolution in vivo imaging of cerebral microvascular flow and structure remains critical to understanding these pathologies. Among them, Ultrasound Localization Microscopy (ULM) allows noninvasive imaging of microvascular network at subwavelength resolution using injected microbubbles, but the approach remains mainly limited to 2D imaging with few volumetric implementations. In this study, we explore in vivo transcranial 3D ULM of the mouse brain using Row-Column Arrays (RCA) and introduce an analysis framework to build a flow-directed vascular graph from the ULM microbubble tracking data, allowing to differentiate between subgraphs of artery-like and vein-like vascular segments. Combined with Allen-based and Radius-based segmentations, we extract metrics including cerebral blood flow (CBF), microbubble (MB) velocity, flowrate, CBV fraction, vascular segment length and tortuosity in sixty-six different vascular classes. This high-sensitivity framework enables in vivo microvascular imaging and quantification in mice and provides a scalable platform for preclinical neurovascular studies in health and disease.

Harmonic Motion Imaging (HMI) is an ultrasound elasticity imaging method that measures the mechanical properties of tissue using amplitude-modulated acoustic radiation force (AM-ARF). By estimating tissues on-axis oscillatory motion, HMI-derived displacement images represent localized relative stiffness and can predict the tumor response to neoadjuvant chemotherapy (NACT) and monitor focused ultrasound (FUS) ablation therapy. Multi-frequency HMI (MF-HMI) excites tissue at various AM frequencies simultaneously, which allows for image optimization without prior knowledge of inclusion size and stiffness. However, challenges remain in size estimation as inconsistent boundary effects result in different perceived sizes across AM frequencies. Herein, we developed an automated tumor and FUS lesion quantification method using a transformer-based multi-modality neural network, HMINet. It was trained on 380 pairs of MF-HMI and B-mode images of phantoms and in vivo orthotopic breast cancer mice (4T1). Test datasets included phantoms (n = 32), in vivo 4T1 mice (n = 24), breast cancer patients (n = 16), and a FUS-induced lesion, with average segmentation accuracy (Dice Similarity Score) of 0.95, 0.86, 0.82, and 0.87, respectively. To increase the generalizability of HMINet, we applied a transfer learning strategy, i.e., fine-tuning the model using patient data. For NACT patients, the displacement ratios (DR) between the tumor and surrounding tissue were calculated based on HMINet-segmented boundaries to predict tumor response based on stiffness changes.