Periodic Reporting for period 1 - Super-FALCON (Super-resolution, ultrafast and deeply-learned contrast ultrasound imaging of the vascular tree.)
Reporting period: 2023-04-01 to 2025-09-30
Healthcare accounts for 10% of the GDP in Europe, of which 15% is spent on cancer and cardio-vascular diseases (CVDs). Cancer and CVDs kill nearly 3 million people per year, which represents 60% of the total death toll and, concurrently with improving treatments, the economic burden continuously intensifies. So far, we do not fully understand the pathogenesis of these diseases and resort either to invasive and time-consuming diagnostic techniques, or to treatment based on generic parameters which disregard patient-specific information. The prime signature of CVDs and cancers is their impact on local vasculature. Namely, these diseases modify blood flow patterns, vessel elasticity and vascular geometry, which current technology cannot quantify with sufficient precision.
Early diagnosis may prevent the disease from becoming symptomatic (e.g. by life-style changes and protective medicines) but this requires a new technology suited to detect and quantify vascular changes at an early stage. To that end, this modality must (1) provide a better-than-10-fold resolution improvement at large depth over current technology (i.e. ~ 50 μm at 10 cm), (2) provide quantitative and relevant information on flow and vasculature, and (3) be fast enough to do so for velocities exceeding 1 m/s (>1000 frames per second).
Super-FALCON proposes a strategy towards a super-resolved (i.e. below diffraction limit), quantitative and real-time technology for imaging flow and vasculature. It is built on contrast-enhanced ultrasound, which is safe, cost-efficient, and widely available, making it a future-proof technology. The core strategy consists in harnessing the nonlinear dynamics of contrast bubbles to image blood flow, from capillaries to arteries, keeping in mind the large differences between both types of vasculatures (dimension, flow velocity, contrast agent response). Achieving this will require new strategies for signal deconvolution and super-resolution, new physical knowledge to describe the behavior of bubbles confined within a viscoelastic capillary vessel, and to be able to compensate for signal distortions induced during propagation. The final goal is to deploy the technique integrating these advances and apply it to an ex vivo liver model that contains a full, multiscale, and perfused vascular tree.
Early diagnosis may prevent the disease from becoming symptomatic (e.g. by life-style changes and protective medicines) but this requires a new technology suited to detect and quantify vascular changes at an early stage. To that end, this modality must (1) provide a better-than-10-fold resolution improvement at large depth over current technology (i.e. ~ 50 μm at 10 cm), (2) provide quantitative and relevant information on flow and vasculature, and (3) be fast enough to do so for velocities exceeding 1 m/s (>1000 frames per second).
Super-FALCON proposes a strategy towards a super-resolved (i.e. below diffraction limit), quantitative and real-time technology for imaging flow and vasculature. It is built on contrast-enhanced ultrasound, which is safe, cost-efficient, and widely available, making it a future-proof technology. The core strategy consists in harnessing the nonlinear dynamics of contrast bubbles to image blood flow, from capillaries to arteries, keeping in mind the large differences between both types of vasculatures (dimension, flow velocity, contrast agent response). Achieving this will require new strategies for signal deconvolution and super-resolution, new physical knowledge to describe the behavior of bubbles confined within a viscoelastic capillary vessel, and to be able to compensate for signal distortions induced during propagation. The final goal is to deploy the technique integrating these advances and apply it to an ex vivo liver model that contains a full, multiscale, and perfused vascular tree.
Super-resolution imaging
- Algorithms:
We have extended our previous 1D network by adding a transducer-element dimension, enabling processing of multi-channel RF data in a single forward pass. The model now outputs super-resolved images directly, rather than deconvolved multi-channel RF. To preserve interpretability and generalizability, we embed differentiable beamforming layers (Delay-and-Sum and Stolt’s f-k migration) directly within the network. These algorithms are implemented in a modular, user-friendly framework allowing non-experts to easily run experiments and compare methods. The multi-channel model successfully deconvolves the lateral direction, overcoming a key limitation of the earlier single-channel model.
We are now investigating realistic 3D pressure fields, with promising preliminary results from the beamforming layer, and are exploring the benefits of hard-coding the beamforming algorithm in terms of performance, dataset size, microbubble density, and invariance to steering angle and transmission. Neural networks have been trained to localize microbubbles in raw ultrasound data for different pulse types, demonstrating that waveform-specific networks outperform others in noisy conditions (see Fig. 5).
Our next step is to generate arbitrary transmit pulses optimized for super-resolution through differentiable simulations.
- Experimental translation:
We have achieved initial experimental validation by applying our algorithms to a monodisperse bubble suspension, yielding encouraging results. To progress from prototype validation to in vitro assessment, we are finalizing a flow phantom that uses hydrodynamic instabilities to create a vortex street as the imaging target. The second version of this setup is now under final testing.
In parallel, we explored particle image velocimetry (PIV) on super-resolved images and demonstrated the critical role of point-spread function (PSF) shape. Preliminary results indicate that 2D super-resolution algorithms significantly reduce the PSF, thereby enhancing velocity estimation accuracy.
Bubbles as capillary mechanical sensors
- Numerical models:
In collaboration with CEMEF (Sophia Antipolis, France), we developed a 3D finite-element model of an ultrasound-driven, coated microbubble confined within a (hyper)elastic capillary. The simulator has been validated against free-field theory and published results. Final refinements are underway, including comparisons with experimental data.
We use this model to study how geometry influences tissue mechanics, bubble dynamics, and wave generation. By varying capillary wall thickness, length, and stiffness, we simulate realistic confinement. Unlike prior studies that neglect surrounding tissue, we find that different wave modes arise depending on wall thickness, significantly affecting microbubble response. Quantifying these effects will improve our understanding of how confinement alters bubble dynamics.
- Experiments:
We characterized the acoustic properties of tissue-mimicking materials in terms of ultrasound attenuation, optical transparency (for high-speed imaging), and stiffness. Based on these results, we developed a multi-capillary phantom embedded in a soft, tunable viscoelastic material with capillary diameters from 15 to 200 μm. The design is being refined to improve perfusion, and we are also developing a nonlinear-propagation compensation method to enhance measurement sensitivity.
Because storage and loss moduli depend strongly on frequency, we recorded ultrasound-driven oscillations of microbubbles in PAA hydrogels using ultra-high-speed imaging and extracted loss and storage moduli from resonance curves. We are extending this approach using a recent theoretical breakthrough on bubble dynamics in gels and atomic-force-microscopy-inspired techniques to determine frequency-dependent viscoelastic properties.
- Algorithms:
We have extended our previous 1D network by adding a transducer-element dimension, enabling processing of multi-channel RF data in a single forward pass. The model now outputs super-resolved images directly, rather than deconvolved multi-channel RF. To preserve interpretability and generalizability, we embed differentiable beamforming layers (Delay-and-Sum and Stolt’s f-k migration) directly within the network. These algorithms are implemented in a modular, user-friendly framework allowing non-experts to easily run experiments and compare methods. The multi-channel model successfully deconvolves the lateral direction, overcoming a key limitation of the earlier single-channel model.
We are now investigating realistic 3D pressure fields, with promising preliminary results from the beamforming layer, and are exploring the benefits of hard-coding the beamforming algorithm in terms of performance, dataset size, microbubble density, and invariance to steering angle and transmission. Neural networks have been trained to localize microbubbles in raw ultrasound data for different pulse types, demonstrating that waveform-specific networks outperform others in noisy conditions (see Fig. 5).
Our next step is to generate arbitrary transmit pulses optimized for super-resolution through differentiable simulations.
- Experimental translation:
We have achieved initial experimental validation by applying our algorithms to a monodisperse bubble suspension, yielding encouraging results. To progress from prototype validation to in vitro assessment, we are finalizing a flow phantom that uses hydrodynamic instabilities to create a vortex street as the imaging target. The second version of this setup is now under final testing.
In parallel, we explored particle image velocimetry (PIV) on super-resolved images and demonstrated the critical role of point-spread function (PSF) shape. Preliminary results indicate that 2D super-resolution algorithms significantly reduce the PSF, thereby enhancing velocity estimation accuracy.
Bubbles as capillary mechanical sensors
- Numerical models:
In collaboration with CEMEF (Sophia Antipolis, France), we developed a 3D finite-element model of an ultrasound-driven, coated microbubble confined within a (hyper)elastic capillary. The simulator has been validated against free-field theory and published results. Final refinements are underway, including comparisons with experimental data.
We use this model to study how geometry influences tissue mechanics, bubble dynamics, and wave generation. By varying capillary wall thickness, length, and stiffness, we simulate realistic confinement. Unlike prior studies that neglect surrounding tissue, we find that different wave modes arise depending on wall thickness, significantly affecting microbubble response. Quantifying these effects will improve our understanding of how confinement alters bubble dynamics.
- Experiments:
We characterized the acoustic properties of tissue-mimicking materials in terms of ultrasound attenuation, optical transparency (for high-speed imaging), and stiffness. Based on these results, we developed a multi-capillary phantom embedded in a soft, tunable viscoelastic material with capillary diameters from 15 to 200 μm. The design is being refined to improve perfusion, and we are also developing a nonlinear-propagation compensation method to enhance measurement sensitivity.
Because storage and loss moduli depend strongly on frequency, we recorded ultrasound-driven oscillations of microbubbles in PAA hydrogels using ultra-high-speed imaging and extracted loss and storage moduli from resonance curves. We are extending this approach using a recent theoretical breakthrough on bubble dynamics in gels and atomic-force-microscopy-inspired techniques to determine frequency-dependent viscoelastic properties.
Potential societal impact:
In the long term, Super-FALCON technology could help diagnose and manage multiple diseases such as fatty liver (25% prevalence), arthritis (24%), type 1 diabetes (9.5%), and Bekhterev’s (1%), all characterized by inflammatory, leaky capillaries. Super-FALCON aims to deliver quantitative, easy-to-interpret data, enabling use by clinicians with minimal training. Its integration into extramural or point-of-care devices could bring high-quality diagnosis directly to patients.
Anticipated scientific outcome:
Super-FALCON is expected to generate over 20 peer-reviewed papers and deepen understanding of confined-bubble dynamics, benefiting biomedical applications (e.g. drug delivery, blood-brain-barrier opening, RF ablation, US-based immunotherapy) and industrial ones (cleaning, catalysis, fire-protective coatings). Once validated, the physics-informed deep-learning approach can extend to other tissue parameters (e.g. nonlinearity coefficient, elastic modulus). WP3 also yields a scatterer-strength map, underscoring Super-FALCON’s broad impact on ultrasound science and beyond.
In the long term, Super-FALCON technology could help diagnose and manage multiple diseases such as fatty liver (25% prevalence), arthritis (24%), type 1 diabetes (9.5%), and Bekhterev’s (1%), all characterized by inflammatory, leaky capillaries. Super-FALCON aims to deliver quantitative, easy-to-interpret data, enabling use by clinicians with minimal training. Its integration into extramural or point-of-care devices could bring high-quality diagnosis directly to patients.
Anticipated scientific outcome:
Super-FALCON is expected to generate over 20 peer-reviewed papers and deepen understanding of confined-bubble dynamics, benefiting biomedical applications (e.g. drug delivery, blood-brain-barrier opening, RF ablation, US-based immunotherapy) and industrial ones (cleaning, catalysis, fire-protective coatings). Once validated, the physics-informed deep-learning approach can extend to other tissue parameters (e.g. nonlinearity coefficient, elastic modulus). WP3 also yields a scatterer-strength map, underscoring Super-FALCON’s broad impact on ultrasound science and beyond.