Protein model quality assessment using rotation-equivariant transformations on point clouds.

Machine learning research concerning protein structure has seen a surge in popularity over the last years with promising advances for basic science and drug discovery. Working with macromolecular structure in a machine learning context requires an adequate numerical representation, and researchers have extensively studied representations such as graphs, discretized 3D grids, and distance maps. As part of CASP14, we explored a new and conceptually simple representation in a blind experiment: atoms as points in 3D, each with associated features. These features-initially just the basic element type of each atom-are updated through a series of neural network layers featuring rotation-equivariant convolutions. Starting from all atoms, we further aggregate information at the level of alpha carbons before making a prediction at the level of the entire protein structure. We find that this approach yields competitive results in protein model quality assessment despite its simplicity and despite the fact that it incorporates minimal prior information and is trained on relatively little data. Its performance and generality are particularly noteworthy in an era where highly complex, customized machine learning methods such as AlphaFold 2 have come to dominate protein structure prediction.

Hierarchical, rotation-equivariant neural networks to select structural models of protein complexes.

Predicting the structure of multi-protein complexes is a grand challenge in biochemistry, with major implications for basic science and drug discovery. Computational structure prediction methods generally leverage predefined structural features to distinguish accurate structural models from less accurate ones. This raises the question of whether it is possible to learn characteristics of accurate models directly from atomic coordinates of protein complexes, with no prior assumptions. Here we introduce a machine learning method that learns directly from the 3D positions of all atoms to identify accurate models of protein complexes, without using any precomputed physics-inspired or statistical terms. Our neural network architecture combines multiple ingredients that together enable end-to-end learning from molecular structures containing tens of thousands of atoms: a point-based representation of atoms, equivariance with respect to rotation and translation, local convolutions, and hierarchical subsampling operations. When used in combination with previously developed scoring functions, our network substantially improves the identification of accurate structural models among a large set of possible models. Our network can also be used to predict the accuracy of a given structural model in absolute terms. The architecture we present is readily applicable to other tasks involving learning on 3D structures of large atomic systems.

Visualizing the mechanical wave of vocal fold tissue during phonation using electroglottogram-triggered ultrasonography.

In order to investigate the vibration pattern, especially the vibrational phase of tissue beneath the vocal fold mucosa, an imaging method called electroglottogram-triggered ultrasonography is proposed. The ultrasonic images of the vocal fold vibration are obtained in the coronal plane from five adult subjects during phonation. The velocity of the vocal fold tissue beneath the mucosal surface is obtained by using a motion estimation method. The results show that the vibration phase difference between tissues at different locations beneath the vocal fold mucosa results in a mechanical wave traveling upward at a speed of 720 to 1826 mm/s.

Visualizing the movement of the contact between vocal folds during vibration by using array-based transmission ultrasonic glottography.

For the purpose of noninvasively visualizing the dynamics of the contact between vibrating vocal fold medial surfaces, an ultrasonic imaging method which is referred to as array-based transmission ultrasonic glottography is proposed. An array of ultrasound transducers is used to detect the ultrasound wave transmitted from one side of the vocal folds to the other side through the small-sized contact between the vocal folds. A passive acoustic mapping method is employed to visualize and locate the contact. The results of the investigation using tissue-mimicking phantoms indicate that it is feasible to use the proposed method to visualize and locate the contact between soft tissues. Furthermore, the proposed method was used for investigating the movement of the contact between the vibrating vocal folds of excised canine larynges. The results indicate that the vertical movement of the contact can be visualized as a vertical movement of a high-intensity stripe in a series of images obtained by using the proposed method. Moreover, a visualization and analysis method, which is referred to as array-based ultrasonic kymography, is presented. The velocity of the vertical movement of the contact, which is estimated from the array-based ultrasonic kymogram, could reach 0.8 m/s during the vocal fold vibration.

Dynamic Coronary Blood Flow Velocity and Wall Shear Stress Estimation Using Ultrasound in an Ex Vivo Porcine Heart.

PURPOSE: Wall shear stress (WSS) is a critically important physical factor contributing to atherosclerosis. Mapping the spatial distribution of local, oscillatory WSS can identify important mechanisms underlying the progression of coronary artery disease. METHODS: In this study, blood flow velocity and time-varying WSS were estimated in the left anterior descending (LAD) coronary artery of an ex vivo beating porcine heart using ultrasound with an 18 MHz linear array transducer aligned with the LAD in a forward-viewing orientation. A pulsatile heart loop with physiologically-accurate flow was created using a pulsatile pump. The coronary artery wall motion was compensated using a local block matching technique. Next, 2D and 3D velocity magnitude and WSS maps in the LAD coronary artery were estimated at different time points in the cardiac cycle using an ultrafast Doppler approach. The blood flow velocity estimated using the presented approach was compared with a commercially-available, calibrated single element blood flow velocity measurement system. RESULTS: The resulting root mean square error (RMSE) of 2D velocity magnitude acquired from a high frequency, linear array transducer was less than 8% of the maximum velocity estimated by the commercial system. CONCLUSION: When implemented in a forward-viewing intravascular ultrasound device, the presented approach will enable dynamic estimation of WSS, an indicator of plaque vulnerability in coronary arteries.

Dirichlet Flow Matching with Applications to DNA Sequence Design.

Discrete diffusion or flow models could enable faster and more controllable sequence generation than autoregressive models. We show that na\"ive linear flow matching on the simplex is insufficient toward this goal since it suffers from discontinuities in the training target and further pathologies. To overcome this, we develop Dirichlet flow matching on the simplex based on mixtures of Dirichlet distributions as probability paths. In this framework, we derive a connection between the mixtures' scores and the flow's vector field that allows for classifier and classifier-free guidance. Further, we provide distilled Dirichlet flow matching, which enables one-step sequence generation with minimal performance hits, resulting in $O(L)$ speedups compared to autoregressive models. On complex DNA sequence generation tasks, we demonstrate superior performance compared to all baselines in distributional metrics and in achieving desired design targets for generated sequences. Finally, we show that our classifier-free guidance approach improves unconditional generation and is effective for generating DNA that satisfies design targets. Code is available at https://github.com/HannesStark/dirichlet-flow-matching.

Equivariant Scalar Fields for Molecular Docking with Fast Fourier Transforms.

Molecular docking is critical to structure-based virtual screening, yet the throughput of such workflows is limited by the expensive optimization of scoring functions involved in most docking algorithms. We explore how machine learning can accelerate this process by learning a scoring function with a functional form that allows for more rapid optimization. Specifically, we define the scoring function to be the cross-correlation of multi-channel ligand and protein scalar fields parameterized by equivariant graph neural networks, enabling rapid optimization over rigid-body degrees of freedom with fast Fourier transforms. The runtime of our approach can be amortized at several levels of abstraction, and is particularly favorable for virtual screening settings with a common binding pocket. We benchmark our scoring functions on two simplified docking-related tasks: decoy pose scoring and rigid conformer docking. Our method attains similar but faster performance on crystal structures compared to the widely-used Vina and Gnina scoring functions, and is more robust on computationally predicted structures. Code is available at https://github.com/bjing2016/scalar-fields.

Effect of skull porosity on ultrasound transmission and wave mode conversion at large incidence angles.

BACKGROUND: Transcranial ultrasound imaging and therapy depend on the efficient transmission of acoustic energy through the skull. Multiple previous studies have concluded that a large incidence angle should be avoided during transcranial-focused ultrasound therapy to ensure transmission through the skull. Alternatively, some other studies have shown that longitudinal-to-shear wave mode conversion might improve transmission through the skull when the incidence angle is increased above the critical angle (i.e., 25° to 30°). PURPOSE: The effect of skull porosity on the transmission of ultrasound through the skull at varying incidence angles was investigated for the first time to elucidate why transmission through the skull at large angles of incidence is decreased in some cases but improved in other cases. METHODS: Transcranial ultrasound transmission at varying incidence angles (0°-50°) was investigated in phantoms and ex vivo skull samples with varying bone porosity (0% to 28.54% ± 3.36%) using both numerical and experimental methods. First, the elastic acoustic wave transmission through the skull was simulated using micro-computed tomography data of ex vivo skull samples. The trans-skull pressure was compared between skull segments having three levels of porosity, that is, low porosity (2.65% ± 0.03%), medium porosity (13.41% ± 0.12%), and high porosity (26.9%). Next, transmission through two 3D-printed resin skull phantoms (compact vs. porous phantoms) was experimentally measured to test the effect of porous microstructure alone on ultrasound transmission through flat plates. Finally, the effect of skull porosity on ultrasound transmission was investigated experimentally by comparing transmission through two ex vivo human skull segments having similar thicknesses but different porosities (13.78% ± 2.05% vs. 28.54% ± 3.36%). RESULTS: Numerical simulations indicated that an increase in transmission pressure occurs at large incidence angles for skull segments having low porosities but not for those with high porosity. In experimental studies, a similar phenomenon was observed. Specifically, for the low porosity skull sample (13.78% ± 2.05%), the normalized pressure was 0.25 when the incidence angle increased to 35°. However, for the high porosity sample (28.54% ± 3.36%), the pressure was no more than 0.1 at large incidence angles. CONCLUSIONS: These results indicate that the skull porosity has an evident effect on the transmission of ultrasound at large incidence angles. The wave mode conversion at large, oblique incidence angles could enhance the transmission of ultrasound through parts of the skull having lower porosity in the trabecular layer. However, for transcranial ultrasound therapy in the presence of highly porous trabecular bone, transmission at a normal incidence angle is preferable relative to oblique incidence angles due to the higher transmission efficiency.

Assessment of kiwifruit firmness by using airflow and laser technique.

Firmness is a valid and widely acknowledged indication of fruit quality that is directly connected to physical structure and mechanical qualities. The deformation signals of kiwifruit for firmness assessment were acquired using an assessment system based on airflow and laser technology in this investigation. Using partial least squares regression (PLSR), genetic algorithm optimization of bp neural network (GA-BP), and an extreme learning machine (ELM), deformation data from kiwifruit was used to create models of Magness-Taylor penetration firmness prediction. The ELM model outperformed the PLSR model, and GA-BP model in the prediction set, with a correlation coefficient of 0.876 and a root mean squared error of 3.576 N in the prediction set. These findings showed that an assessment system based on airflow and laser techniques can be utilized to assess the firmness of kiwifruit quickly and nondestructively.

A Transverse Velocity Spectral Estimation Method for Ultrafast Ultrasound Doppler Imaging.

A novel transverse velocity spectral estimation method is proposed to estimate the velocity component in the direction transverse to the beam axis for ultrafast imaging. The transverse oscillation was introduced by filtering the envelope data after the axial oscillation was removed. The complex transverse oscillated signal was then used to estimate the transverse velocity spectrum and mean velocity. In simulations, both steady flow with a parabolic flow profile and temporally varying flow were simulated to investigate the performance of the proposed method. Next, the proposed approach was used to estimate the flow velocity in a phantom with pulsatile flow, and finally, this method was applied in vivo in a small animal model. Results of the simulation study indicate that the proposed method provided an accurate velocity spectrogram for beam-to-flow angles from 45° to 90°, without significant performance degradation as the angle decreased. For the simulation of temporally varying flow, the proposed method had a reduced bias ( % versus 73.3%) and higher peak-to-background ratio (PBR) (>15.6 versus 10.5 dB) compared to previous methods. Results in a vessel phantom show that the temporally varying flow velocity can be estimated in the transverse direction obtained using the spectrogram produced by the proposed method operating on the envelope data. Finally, the proposed method was used to map the microvascular blood flow velocity in the mouse spinal cord, demonstrating the estimation of pulsatile blood flow in both the axial and transverse directions in vivo over several cardiac cycles.

EigenFold: Generative Protein Structure Prediction with Diffusion Models.

Protein structure prediction has reached revolutionary levels of accuracy on single structures, yet distributional modeling paradigms are needed to capture the conformational ensembles and flexibility that underlie biological function. Towards this goal, we develop EigenFold, a diffusion generative modeling framework for sampling a distribution of structures from a given protein sequence. We define a diffusion process that models the structure as a system of harmonic oscillators and which naturally induces a cascading-resolution generative process along the eigenmodes of the system. On recent CAMEO targets, EigenFold achieves a median TMScore of 0.84, while providing a more comprehensive picture of model uncertainty via the ensemble of sampled structures relative to existing methods. We then assess EigenFold's ability to model and predict conformational heterogeneity for fold-switching proteins and ligand-induced conformational change. Code is available at https://github.com/bjing2016/EigenFold.

Very Low Frequency Radial Modulation for Deep Penetration Contrast-Enhanced Ultrasound Imaging.

Contrast-enhanced ultrasound imaging allows vascular imaging in a variety of diseases. Radial modulation imaging is a contrast agent-specific imaging approach for improving microbubble detection at high imaging frequencies (≥7.5 MHz), with imaging depth limited to a few centimeters. To provide high-sensitivity contrast-enhanced ultrasound imaging at high penetration depths, a new radial modulation imaging strategy using a very low frequency (100 kHz) ultrasound modulation wave in combination with imaging pulses ≤5 MHz is proposed. Microbubbles driven at 100 kHz were imaged in 10 successive oscillation states by manipulating the pulse repetition frequency to unlock the frame rate from the number of oscillation states. Tissue background was suppressed using frequency domain radial modulation imaging (F-RMI) and singular value decomposition-based radial modulation imaging (S-RMI). One hundred-kilohertz modulation resulted in significantly higher microbubble signal magnitude (63-88 dB) at the modulation frequency relative to that without 100-kHz modulation (51-59 dB). F-RMI produced images with high contrast-to-tissue ratios (CTRs) of 15 to 22 dB in a stationary tissue phantom, while S-RMI further improved the CTR (19-26 dB). These CTR values were significantly higher than that of amplitude modulation pulse inversion images (11.9 dB). In the presence of tissue motion (1 and 10 mm/s), S-RMI produced high-contrast images with CTR up to 18 dB; however, F-RMI resulted in minimal contrast enhancement in the presence of tissue motion. Finally, in transcranial ultrasound imaging studies through a highly attenuating ex vivo cranial bone, CTR values with S-RMI were as high as 23 dB. The proposed technique demonstrates successful modulation of microbubble response at 100 kHz for the first time. The presented S-RMI low-frequency radial modulation imaging strategy represents the first demonstration of real-time (20 frames/s), high-penetration-depth radial modulation imaging for contrast-enhanced ultrasound imaging.

Effect of Skull Porous Trabecular Structure on Transcranial Ultrasound Imaging in the Presence of Elastic Wave Mode Conversion at Varying Incidence Angle.

With the advancement of aberration correction techniques, transcranial ultrasound imaging has exhibited great potential in applications such as imaging neurological function and guiding therapeutic ultrasound. However, the feasibility of transcranial imaging varies among individuals because of the differences in skull acoustic properties. To better understand the fundamental mechanisms underlying the variation in imaging performance, the effect of the structure of the porous trabecular bone on transcranial imaging performance (i.e., target localization errors and resolution) was investigated for the first time through the use of elastic wave simulations and experiments. Simulation studies using high-resolution computed tomography data from ex vivo skull samples revealed that imaging at large incidence angles reduced the target localization error for skulls having low porosity; however, as skull porosity increased, large angles of incidence resulted in degradation of resolution and increased target localization errors. Experimental results indicate that imaging at normal incidence introduced a localization error of 1.85 ± 0.10 mm, while imaging at a large incidence angle (40°) resulted in an increased localization error of 6.54 ± 1.33 mm and caused a single point target to no longer appear as a single, coherent target in the resulting image, which is consistent with simulation results. This first investigation of the effects of skull microstructure on transcranial ultrasound imaging indicates that imaging performance is highly dependent on the porosity of the skull, particularly at non-normal angles of incidence.

Forward-viewing estimation of 3D blood flow velocity fields by intravascular ultrasound: Influence of the catheter on velocity estimation in stenoses.

Coronary artery disease is the most common type of cardiovascular disease, affecting > 18 million adults, and is responsible for > 365 k deaths per year in the U.S. alone. Wall shear stress (WSS) is an emerging indicator of likelihood of plaque rupture in coronary artery disease, however, non-invasive estimation of 3-D blood flow velocity and WSS is challenging due to the requirement for high spatial resolution at deep penetration depths in the presence of significant cardiac motion. Thus we propose minimally-invasive imaging with a catheter-based, 3-D intravascular forward-viewing ultrasound (FV US) transducer and present experiments to quantify the effect of the catheter on flow disturbance in stenotic vessel phantoms with realistic velocities and luminal diameters for both peripheral (6.33 mm) and coronary (4.74 mm) arteries. An external linear array ultrasound transducer was used to quantify 2-D velocity fields in vessel phantoms under various conditions of catheter geometry, luminal diameter, and position of the catheter relative to the stenosis at a frame rate of 5000 frames per second via a particle imaging velocimetry (PIV) approach. While a solid catheter introduced an underestimation of velocity measurement by > 20% relative to the case without a catheter, the hollow catheter introduced < 10% velocity overestimation, indicating that a hollow catheter design allowing internal blood flow reduces hemodynamic disturbance. In addition, for both peripheral and coronary arteries, the hollow catheter introduced < 3% deviation in flow velocity at the minimum luminal area compared to the control case. Finally, an initial comparison was made between velocity measurements acquired using a low frequency, catheter-based, 3-D intravascular FV US transducer and external linear array measurements, with relative error < 12% throughout the region of interest for a flow rate of 150 mL/min. While further system development is required, results suggest intravascular ultrasound characterization of blood flow velocity fields in stenotic vessels could be feasible with appropriate catheter design.

Patient-Specific 3D Bioprinted Models of Developing Human Heart.

The heart is the first organ to develop in the human embryo through a series of complex chronological processes, many of which critically rely on the interplay between cells and the dynamic microenvironment. Tight spatiotemporal regulation of these interactions is key in heart development and diseases. Due to suboptimal experimental models, however, little is known about the role of microenvironmental cues in the heart development. This study investigates the use of 3D bioprinting and perfusion bioreactor technologies to create bioartificial constructs that can serve as high-fidelity models of the developing human heart. Bioprinted hydrogel-based, anatomically accurate models of the human embryonic heart tube (e-HT, day 22) and fetal left ventricle (f-LV, week 33) are perfused and analyzed both computationally and experimentally using ultrasound and magnetic resonance imaging. Results demonstrate comparable flow hemodynamic patterns within the 3D space. We demonstrate endothelial cell growth and function within the bioprinted e-HT and f-LV constructs, which varied significantly in varying cardiac geometries and flow. This study introduces the first generation of anatomically accurate, 3D functional models of developing human heart. This platform enables precise tuning of microenvironmental factors, such as flow and geometry, thus allowing the study of normal developmental processes and underlying diseases.

A 3D Bioprinted In Vitro Model of Pulmonary Artery Atresia to Evaluate Endothelial Cell Response to Microenvironment.

Vascular atresia are often treated via transcatheter recanalization or surgical vascular anastomosis due to congenital malformations or coronary occlusions. The cellular response to vascular anastomosis or recanalization is, however, largely unknown and current techniques rely on restoration rather than optimization of flow into the atretic arteries. An improved understanding of cellular response post anastomosis may result in reduced restenosis. Here, an in vitro platform is used to model anastomosis in pulmonary arteries (PAs) and for procedural planning to reduce vascular restenosis. Bifurcated PAs are bioprinted within 3D hydrogel constructs to simulate a reestablished intervascular connection. The PA models are seeded with human endothelial cells and perfused at physiological flow rate to form endothelium. Particle image velocimetry and computational fluid dynamics modeling show close agreement in quantifying flow velocity and wall shear stress within the bioprinted arteries. These data are used to identify regions with greatest levels of shear stress alterations, prone to stenosis. Vascular geometry and flow hemodynamics significantly affect endothelial cell viability, proliferation, alignment, microcapillary formation, and metabolic bioprofiles. These integrated in vitro-in silico methods establish a unique platform to study complex cardiovascular diseases and can lead to direct clinical improvements in surgical planning for diseases of disturbed flow.

Phase Modulation Beamforming for Ultrafast Plane-Wave Imaging.

In order to improve the spatial resolution for high-frame-rate imaging, a new image formation approach is proposed based on introducing very weak aberration into received data, then combining the multiple results by taking the pixel-wise standard deviation of multiple aberrated images and subtracting the result from the delay-and-sum image. This approach is demonstrated in simulations, tissue-mimicking phantom experiments, and in vivo imaging. Simulations indicate the lateral full-width half-maximum (FWHM) of targets decreases by 38.24% ± 6.38%. In imaging wire targets in a tissue-mimicking phantom at 7.8 MHz, wire target FWHM decreases by 35.91% ± 5.39%. However, contrast was observed to decrease by 1.23 dB and contrast-to-noise ratio (CNR) by 18.5% in phantom studies due to the subtraction of similar images, which increases the number of dark pixels in the image. Finally, the proposed technique is tested in vivo, with images showing improvements similar to those in tissue-mimicking phantoms, including increased separation between closely spaced targets.

Transcranial activation and imaging of low boiling point phase-change contrast agents through the temporal bone using an ultrafast interframe activation ultrasound sequence.

PURPOSE: As a cavitation enhancer, low boiling point phase-change contrast agents (PCCA) offer potential for ultrasound-mediated drug delivery in applications including intracerebral hemorrhage or brain tumors. In addition to introducing cavitation, ultrasound imaging also has the ability to provide guidance and monitoring of the therapeutic process by localizing delivery events. However, the highly attenuating skull poses a challenge for achieving an image with useful contrast. In this study, the feasibility of transcranial activation and imaging of low boiling point PCCAs through the human temporal bone was investigated by using a low frequency ultrafast interframe activation ultrasound (UIAU) imaging sequence with singular value decomposition-based denoising. METHODS: Lipid-shelled PCCAs filled with decafluorobutane were activated and imaged at 37°C in tissue-mimicking phantoms both without and with an ex vivo human skull using the new UIAU sequence and a low frequency diagnostic transducer array at frequencies from 1.5 to 3.5 MHz. A singular value decomposition-based denoising filter was developed and used to further enhance transcranial image contrast. The contrast-to-tissue ratio (CTR) and contrast enhancement (CE) of UIAU was quantitatively evaluated and compared with the amplitude modulation pulse inversion (AMPI) and vaporization detection imaging (VDI) approaches. RESULTS: Image results demonstrate enhanced contrast in the phantom channel with suppressed background when the low boiling point PCCA was activated both without and with an ex vivo human skull using the UIAU sequence. Quantitative results show that without the skull, low frequency UIAU imaging provided significantly higher image contrast (CTR ≥ 18.56 dB and CE ≥ 18.66 dB) than that of AMPI and VDI (P < 0.05). Transcranial imaging results indicated that the CE of UIAU (≥18.80 dB) was significantly higher than AMPI for free-field activation pressures of 5 and 6 MPa. The CE of UIAU is also significantly higher than that of VDI when PCCAs were activated at 2.5 MHz and 3 MHz (P < 0.05). The CTR (23.30 [20.07-25.56] dB) of denoised UIAU increased by 12.58 dB relative to the non-denoised case and was significantly higher than that of AMPI at an activation pressure of 4 MPa (P < 0.05). CONCLUSIONS: Results indicate that low boiling point PCCAs can be activated and imaged at low frequencies including imaging through the temporal bone using the UIAU sequence. The UIAU imaging approach provides higher contrast than AMPI and VDI, especially at lower activation pressures with additional removal of electronic noise.

High contrast power Doppler imaging in side-viewing intravascular ultrasound imaging via angular compounding.

The ability to assess likelihood of plaque rupture can determine the course of treatment in coronary artery disease. One indicator of plaque vulnerability is the development of blood vessels within the plaque, or intraplaque neovascularization. In order to visualize these vessels with increased sensitivity in the cardiac catheterization lab, a new approach for imaging blood flow in small vessels using side-viewing intravascular ultrasound (IVUS) is proposed. This approach based on compounding adjacent angular acquisitions was evaluated in tissue mimicking phantoms and ex vivo vessels. In phantom studies, the Doppler CNR increased from 3.3 ± 1.0 to 13 ± 2.6 (conventional clutter filtering) and from 1.9 ± 0.15 to 7.5 ± 1.1 (SVD filtering) as a result of applying angular compounding. When imaging flow at a rate of 5.6 mm/s in 200 µm tubes adjacent to the lumen of ex vivo porcine arteries, the Doppler CNR increased from 5.3 ± 0.95 to 7.2 ± 1.3 (conventional filtering) and from 23 ± 3.3 to 32 ± 6.7 (SVD filtering). Applying these strategies could allow increased sensitivity to slow flow in side-viewing intravascular ultrasound imaging.

Imaging the Activation of Low-Boiling-Point Phase-Change Contrast Agents in the Presence of Tissue Motion Using Ultrafast Inter-frame Activation Ultrasound Imaging.

Nanoscale phase-change contrast agents (PCCAs) have been found to have great potential in non-invasive extravascular imaging and therapeutic delivery. However, the contrast-to-tissue ratio (CTR) of PCCA images is usually limited because of either physiological motion or incomplete cancelation of tissue signal. Therefore, to improve the CTR of PCCA images in the presence of physiological motion, a new imaging technique, ultrafast inter-frame activation ultrasound (UIAU) imaging, is proposed and validated. Results of studies with controlled motion in tissue-mimicking phantoms indicate UIAU could provide significantly higher CTRs (maximum: 17.3 ± 0.9 dB) relative to conventional pulse inversion imaging (maximum CTR: 3.4 ± 1.4 dB). UIAU has CTRs up to 16.1 ± 1.0 dB relative to 3.9 ± 2.3 dB for differential imaging in the presence of physiological motion at 20 mm/s. In vivo imaging of PCCAs in the rat liver also reveals the ability of UIAU to enhance PCCA image contrast in the presence of physiological motion.