Towards Modeling the Hemodynamic Impact of Mitral and Aortic Valve Repair in Patients with Left Ventricular Assist Devices
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Physics. The day in general physics: instruments, methods and results across the field.
Kai: We're looking at the paper: "Towards Modeling the Hemodynamic Impact of Mitral and Aortic Valve Repair in Patients with Left Ventricular Assist Devices".
Mira: Here's the gist of the paper. Towards Modeling the Hemodynamic Impact of Mitral and Aortic Valve Repair in Patients with Left Ventricular Assist Devices is a patient-specific computational study that evaluates how simulated mitral and aortic valve repairs influence left heart hemodynamics in patients supported by a Left Ventricular Assist Device (LVAD).
Kai: The simulations suggest that valve repair improved cardiac output, reduced pulmonary congestion, and enhanced right ventricular loading conditions Modeling Framework The study applies a patient-specific, image-based computational modeling framework to evaluate the hemodynamic impact of simulated mitral and aortic valve repair in five LVAD-supported patients Each patient was modeled under four conditions: (patient-specific LVAD-supported state), simulated mitral valve (MV) repair, simulated aortic valve (AV) repair, and simulated combined MV&AV repair The models integrate dynamic CT imaging, echocardiography, catheterization data, and device-specific LVAD parameters into a coupled 3D–0D simulation pipeline Valve dynamics are governed by transvalvular pressure and flow, allowing physiological modeling of regurgitant lesions and surgical repair Patient Data Characteristics The five patients exhibit substantial heterogeneity in both cardiac structure and hemodynamic state Ventricular volumes vary markedly across cases, with end-diastolic volume indices (EDVI) ranging from 77.8 to 313.5 mL/m2 and end-systolic volume indices (ESVI) ranging from 64 to 295 mL/m2 Prior to LVAD implantation, all patients had at least moderate MR The severity of regurgitation and insufficiency was determined based on an echocardiogram (TTE) exam Variability is also observed in global hemodynamic measures, including cardiac output, pulmonary pressures, and right ventricular loading conditions Simulated Mitral Valve Repair In silico mitral valve repair yielded physiologically consistent hemodynamic improvements across all five patient-specific models On average, the forward cardiac output increased 21.2±15.91% (p = 0.02) The reduction in mitral regurgitation was consistently associated with lower left atrial and pulmonary venous pressures, along with elevated systolic pressures in the left ventricle following simulated repair During systole, all models demonstrated left ventricular pressures exceeding those in the aorta, resulting in physiological opening of the aortic valve and forward ejection In silico mitral valve repair also reduced left atrial flow disturbances and pulmonary venous congestion, leading to an 2.5%-45.1% reduction in peak RV pressure and an 0.4%-34.7% decrease in RV end-diastolic volume On average, the RVEF and PAPi increased by 21.9±16.42% (p = 0.02) and 24.5±23.55% (p = 0.04) Simulated Aortic Valve Repair In all cases, in silico AV repair (n=4) removed the high backflow into the LV at the aortic valve Simulated aortic valve repair did not improve aortic valve opening or forward flow in any case; however, it produced a modest reduction in mitral regurgitant volume (21.5% decrease) compared to the patient-specific state The intervention also lowered left ventricular filling pressures, with mean LV end-diastolic pressure decreasing on average 33.6±11.68% (p = 0.005) following in silico repair When evaluating RV metrics, simulated aortic valve repair yielded similar trends to simulated MV repair cases Combined Valve Repair Effects When both valves were repaired simultaneously, the hemodynamic improvements were more pronounced Averaged across four patients (A–D), forward cardiac output increased 33.6±9.72% (p = 0.011) Combined in silico mitral and aortic valve repair lowered the LV diastolic pressure more than isolated MV or AV in silico repair Combined mitral and aortic valve in silico repair also reduced pulmonary pressures, which may lead to improved RV function The average RVEF, PAPi, and Ees/Ea ratio increased by 33.8±3.02% (p = 0.008), 42.2±12.89% (p = 0.011), and 58.2±25.81% (p = 0.022) respectively Residence Time Analysis The relative residence time (rRT) for Patient B is shown in Figure 8 Following simulated mitral valve repair, aortic valve opening improved washout within the aorta Combined MV&AV repair produced the largest and most consistent reductions across compartments, particularly in the LA (up to ∼37%) and LV (up to ∼11%), indicating synergistic improvements in washout when both valves are addressed Correcting mitral and aortic insufficiency could not only improves left heart hemodynamics and washout efficiency but may also mitigate progressive right ventricular dysfunction in the LVAD population These findings highlight that correcting regurgitant lesions in LVAD patients may prevent progressive RV–PA uncoupling and improve RV loading conditions The patient-specific nature of simulated repair outcomes is evident across the cohort The small sample size (N = 5) limits the ability to draw definitive statistical conclusions Limitations The computational framework does not capture dynamic or long-term cardiovascular adaptation Valve function was modeled using simplified planar surfaces with prescribed effective orifice areas rather than fully resolved leaflet anatomy The model does not account for perioperative factors that may influence actual valve repair outcomes, such as surgical technique, annuloplasty ring sizing, residual leaflet tethering, or acute changes in ventricular loading following surgery Ethics All data collection and analysis were conducted under IRB-approved protocol HUM00196629 (approved April 2021) Acknowledgments This work was supported by the National Institute of Health National Heart, Lung, & Blood Institute [No.
Paper summary: Kai: So we're looking at a paper called "Towards Modeling the Hemodynamic Impact of Mitral and Aortic Valve Repair in Patients with Left Ventricular Assist Devices." This study looks at how fixing those heart valves affects blood flow when a patient is on an LVAD.
Mira: Yeah, the core idea here is using patient-specific computational models to see what happens hemodynamically when you simulate repairing both the mitral and aortic valves in these patients. It claims that simulated valve repair improves cardiac output, lowers pulmonary congestion, and actually enhances the loading conditions on the right ventricle.
Lev: I mean, it’s about taking this complex physiological situation—an LVAD patient with pre-existing valve problems—and testing different surgical scenarios in a computer simulation to see the expected effect on blood dynamics. It sets up these four conditions: the patient's actual state, simulated mitral repair, simulated aortic repair, and combined mitral and aortic repair.
Kai: Exactly. They used dynamic CT imaging, echocardiography data, catheterization data, and device-specific parameters to build these coupled three dimensional zero dimensional simulations. The valve dynamics are governed by transvalvular pressure and flow so they can model regurgitant lesions or surgical repairs physiologically.
Mira: The patient group itself is pretty varied. They looked at five LVAD-supported patients who had a lot of differences in their cardiac structure and hemodynamic state, with end diastolic volume indices ranging from seventy-seven point eight to three hundred thirteen point five milliliters per square meter.
Lev: And they noted that before the LVAD went in, all these patients already had at least moderate mitral regurgitation. The severity of that regurgitation and insufficiency was determined using a transthoracic echocardiogram exam, and variability in things like cardiac output and pulmonary pressures was also observed.
Kai: They tested the simulated mitral valve repair across all five patient-specific models, and they found some consistent results there too. On average, the forward cardiac output increased by twenty-one point two plus or minus fifteen point nine one percent (p equals zero point zero two) <ref:2609.26939#pg3>.
Mira: That increase in cardiac output came with a reduction in mitral regurgitation, which was consistently linked to lower left atrial and pulmonary venous pressures, along with elevated systolic pressures in the left ventricle after the simulated repair.
Paper summary: Lev: During systole, all those models showed left ventricular pressures exceeding those in the aorta, which resulted in a physiological opening of the aortic valve and forward ejection <ref:2609.26939#pg3>. It’s interesting that they're modeling this dynamic interaction between the chambers.
Kai: It also reduced left atrial flow disturbances and pulmonary venous congestion, leading to a reduction of two point five percent to forty-five point one percent in peak right ventricular pressure and a decrease of zero point four percent to thirty-four point seven percent in right ventricular end diastolic volume <ref:2609.26939#pg3>.
Mira: On average, the rate-volume efficiency fraction and pulmonary artery pressure index increased by twenty-one point nine plus or minus sixteen point four two percent (p equals zero point zero two) and twenty-four point five plus or minus twenty-three point five five percent (p equals zero point zero four) <ref:2609.26939#pg3>.
Lev: Now, they also looked at the simulated aortic valve repair. They found that in all cases, simulated aortic valve repair did not improve aortic valve opening or forward flow in any case, but it did produce a modest reduction in mitral regurgitant volume of twenty-one point five percent compared to the patient-specific state.
Kai: The intervention also lowered left ventricular filling pressures, with the mean left ventricular end diastolic pressure decreasing on average by thirty-three point six plus or minus eleven point sixty-eight percent (p equals zero point zero zero five) after the simulated repair <ref:2609.26939#pg3>.
Mira: When evaluating the right ventricle metrics, simulated aortic valve repair showed similar trends to the cases where they did the simulated mitral valve repair.
Lev: So when they looked at combining both repairs simultaneously, things got a bit more pronounced hemodynamically. Averaged across four patients from A to D, forward cardiac output increased by thirty-three point six plus or minus nine point seventy-two percent (p equals zero point zero one) <ref:2609.26939#pg3>.
Kai: Combined mitral and aortic valve repair lowered the left ventricular diastolic pressure more than either the isolated mitral or aortic repair did by itself. And they also found that combined repairs reduced pulmonary pressures, which could lead to improved right ventricular function <ref:2609.26939#pg3>.
Paper summary: Mira: The average rate-volume efficiency fraction, pulmonary artery pressure index, and end systolic ejection ratio all increased by thirty-three point eight plus or minus three point two percent (p equals zero point zero zero eight), forty-two point two plus or minus twelve point eighty-nine percent (p equals zero point zero eleven), and fifty-eight point two plus or minus twenty-five point eighty one percent (p equals zero point zero two two) <ref:2609.26939#pg3>.
Lev: They also looked at the relative residence time for Patient B, which showed that after simulated mitral valve repair, aortic valve opening improved washout within the aorta. When they did both repairs together, it produced the largest and most consistent reductions across compartments, particularly in the left atrium up to about thirty-seven percent and in the left ventricle up to about eleven percent.
Kai: That suggests a synergistic effect when you address both valves because of how it affects washout efficiency. It points toward a real improvement in how blood moves through the system, not just static pressures.
Mira: So, looking at the whole picture of this paper, "Towards Modeling the Hemodynamic Impact of Mitral and Aortic Valve Repair in Patients with Left Ventricular Assist Devices," it really shows that correcting mitral and aortic insufficiency could improve left heart hemodynamics and washout efficiency.
Lev: And they suggest that doing this might also mitigate progressive right ventricular dysfunction in the LVAD population. It implies that fixing regurgitant lesions in these patients might prevent progressive right ventricle pulmonary artery uncoupling and actually improve loading conditions for the right ventricle.
Kai: The paper highlights how patient-specific the repair outcomes are across this cohort, even though it's a small sample size of five patients, which limits how definitive they can be with statistics.
Mira: So, what does this mean in practice? It shows that addressing regurgitation in LVAD patients could potentially help manage the right heart performance over time.
Lev: The limitation they state is that their computational framework doesn't capture dynamic or long-term cardiovascular adaptation, and valve function was modeled using simplified planar surfaces with prescribed effective orifice areas instead of fully resolved leaflet anatomy.
Conclusion: Kai: So this paper is about using computer models to see how fixing mitral and aortic valves affects the blood flow in people who have these left ventricular assist devices, or LVADs.
Mira: Right, it's trying to figure out if fixing those valves can actually improve things like cardiac output and reduce congestion in the lungs.
Lev: The authors are using patient-specific modeling based on imaging data from five different patients to test what happens when they simulate repairs—mitral repair, aortic repair, or both.
Kai: They found that repairing the mitral valve alone gave a solid lift in cardiac output and helped lower pressures in the lungs and left atrium.
Mira: That's interesting because it shows that fixing the mitral issue has a pretty consistent effect across all five patients they studied.
Lev: But they also looked at aortic repair by itself, which didn't really change the forward flow much, though it did lower the filling pressures in the left ventricle.
Kai: When you look at both repairs done together, that’s where things get more dramatic—the cardiac output jumped up significantly and pulmonary pressures dropped further.
Mira: That combined effect suggests that addressing both valve issues simultaneously could be really beneficial for improving right heart function overall in these LVAD patients.
Lev: It points toward a mechanism where fixing those leaks might stop the right ventricle from getting worse over time, preventing some of that progressive dysfunction they see.
Kai: So, the authors are suggesting that correcting these regurgitant lesions isn't just about the left side; it could actually stabilize the whole right heart system in these patients.
Mira: It’s a big picture idea because it connects valve repair directly to managing a key problem—right ventricle loading conditions—in people who rely on these devices.
Lev: They did have to acknowledge that since they only looked at five patients, the statistical conclusions are limited, and their model uses simplified surfaces instead of the actual complex leaflet shapes.
Kai: It means this is a strong simulation showing a potential path forward, but we still need more data and more detailed models before we can say it's a definitive solution for every single patient.
Mira: So the next thing to look at is how these findings might translate into actual clinical practice, especially since they showed that combining both repairs yields the biggest improvement in washout efficiency.
Mia Bonini, Michael Ferguson, Marc Hirshvogel, Maximilian Balmus, Paul C. Tang, Francis Pagani, David Nordsletten
Department of Biomedical Engineering, University of Michigan Department of Cardiac Surgery Mayo Clinic
physics.med-ph, physics.comp-ph
Submitted: 2026-09-22
Updated: 2026-09-22
Journal ref: M. Bonini et al., Towards modeling the hemodynamic impact of mitral and aortic valve repair in patients with left ventricular assist devices, Comput. Biol. Med., vol. 217, no. 112014, p. 112014, Nov. 2026
DOI: 10.1016/j.compbiomed.2026.112014
License: http://creativecommons.org/licenses/by-nc-nd/4.0/
Importance score: 80/100
The gist: Towards Modeling the Hemodynamic Impact of Mitral and Aortic Valve Repair in Patients with Left Ventricular Assist Devices is a patient-specific computational study that evaluates how simulated…
Key concepts
- Patient-specific computational modeling
- This involves creating detailed computer models based on an individual patient's actual medical images (like CT scans). Instead of using generic models, researchers use these personalized digital twins to simulate how specific surgeries, such as valve repairs, will impact that unique patient's heart and blood flow.
- Left Ventricular Assist Device (LVAD)
- An LVAD is a mechanical device implanted in the heart to help pump blood. This study focuses on patients already using an LVAD. The goal is to see how fixing the valves in these patients changes how effectively their heart and the device work together to manage blood flow.
- Hemodynamic impact
- This refers to how changes in blood flow (pressure, volume, and speed) within the heart affect its function. The study measured things like cardiac output (how much blood is pumped per minute) and pulmonary pressures to quantify these changes after simulating valve repairs.
Terminology
Summary
Towards Modeling the Hemodynamic Impact of Mitral and Aortic Valve Repair in Patients with Left Ventricular Assist Devices is a patient-specific computational study that evaluates how simulated mitral and aortic valve repairs influence left heart hemodynamics in patients supported by a Left Ventricular Assist Device (LVAD). p. 0 Highlights
Towards Modeling the Hemodynamic Impact of Mitral and Aortic Valve Repair in
Patients with Left Ventricular Assist Devices
Mia Bonini, Michael Ferguson, Marc Hirshvogel, Maximilian Balmus, Paul C. Tang, Francis Pagani,
David Nordsletten
• Patient-specific 3D–0D hemodynamic models simulating aortic and mitral valve repair effects
in LVAD hearts
• simulated mitral repair improves aortic valve opening under LVAD support
• simulated aortic and mitral valve repair improves right ventricle loading conditions
• Residence time analysis reveals reduced aortic root stasis after simulated mitral valve repair
arXiv:2609.Where the paper says it
Modeling Framework
The study applies a patient-specific, image-based computational modeling framework to evaluate the hemodynamic impact of simulated mitral and aortic valve repair in five LVAD-supported patients Each patient was modeled under four conditions: (patient-specific LVAD-supported state), simulated mitral valve (MV) repair, simulated aortic valve (AV) repair, and simulated combined MV&AV repair The models integrate dynamic CT imaging, echocardiography, catheterization data, and device-specific LVAD parameters into a coupled 3D–0D simulation pipeline Valve dynamics are governed by transvalvular pressure and flow, allowing physiological modeling of regurgitant lesions and surgical repair
Patient Data Characteristics
The five patients exhibit substantial heterogeneity in both cardiac structure and hemodynamic state Ventricular volumes vary markedly across cases, with end-diastolic volume indices (EDVI) ranging from 77.8 to 313.5 mL/m2 and end-systolic volume indices (ESVI) ranging from 64 to 295 mL/m2 Prior to LVAD implantation, all patients had at least moderate MR The severity of regurgitation and insufficiency was determined based on an echocardiogram (TTE) exam Variability is also observed in global hemodynamic measures, including cardiac output, pulmonary pressures, and right ventricular loading conditions
Simulated Mitral Valve Repair
In silico mitral valve repair yielded physiologically consistent hemodynamic improvements across all five patient-specific models On average, the forward cardiac output increased 21.2±15.91% (p = 0.02) The reduction in mitral regurgitation was consistently associated with lower left atrial and pulmonary venous pressures, along with elevated systolic pressures in the left ventricle following simulated repair During systole, all models demonstrated left ventricular pressures exceeding those in the aorta, resulting in physiological opening of the aortic valve and forward ejection In silico mitral valve repair also reduced left atrial flow disturbances and pulmonary venous congestion, leading to an 2. p. 1 The simulations suggest that valve repair improved cardiac output, reduced pulmonary
congestion, and enhanced right ventricular loading conditions. Notably, mitral valve repair restored
aortic valve opening during the cardiac cycle, which improved sinus washout and reduced blood
residence time within the aortic root—factors associated with lower thrombotic risk. The models integrate dynamic
CT imaging, echocardiography, catheterization data, and device-specific LVAD parameters into a
coupled 3D–0D simulation pipeline. Valve dynamics are governed by transvalvular pressure and
flow, allowing physiological modeling of regurgitant lesions and surgical repair. The five patients considered in this study exhibit substantial heterogeneity in both cardiac
structure and hemodynamic state (see Table 1). Ventricular volumes vary markedly across cases,
with end-diastolic volume indices (EDVI) ranging from 77.Where the paper says it
Where the paper says it
Where the paper says it
Simulated Aortic Valve Repair
In all cases, in silico AV repair (n=4) removed the high backflow into the LV at the aortic valve Simulated aortic valve repair did not improve aortic valve opening or forward flow in any case; however, it produced a modest reduction in mitral regurgitant volume (21.5% decrease) compared to the patient-specific state The intervention also lowered left ventricular filling pressures, with mean LV end-diastolic pressure decreasing on average 33. p. 4 Abbreviations:
BSA, body surface area; LV, left ventricle; ESVI, end systolic volume index; EDVI, end diastolic volume index;
Abbreviations: CO, cardiac output; sBP, systolic blood pressure; dBP, diastolic blood pressure; RAP, right atrial
pressure; RVP, right ventricle pressure; PAP, pulmonary artery pressure;PCWP, pulmonary capillary wedge pressure;
RVEDP, right ventricle end diastolic pressure; RF, regurgitant fraction; RVEDV, right ventricle end diastolic volume;
Vol, Volume; mod, moderate; sev, severe. Anatomic Reconstruction and 3D Model Generation
Contrast-enhanced axial CT scans at end diastole were used to reconstruct patient-specific
anatomy.Where the paper says it
Combined Valve Repair Effects
When both valves were repaired simultaneously, the hemodynamic improvements were more pronounced Averaged across four patients (A–D), forward cardiac output increased 33.6±9.72% (p = 0.011) Combined in silico mitral and aortic valve repair lowered the LV diastolic pressure more than isolated MV or AV in silico repair Combined mitral and aortic valve in silico repair also reduced pulmonary pressures, which may lead to improved RV function The average RVEF, PAPi, and Ees/Ea ratio increased by 33. p. 1 Each patient was modeled
under four conditions: (patient-specific LVAD-supported state), simulated mitral valve (MV) repair,
simulated aortic valve (AV) repair, and simulated combined MV&AV repair. Because validation
data for valve repair were unavailable, the simulated repair scenarios are exploratory in silico
interventions based on clinically validated patient-specific models.Where the paper says it
Residence Time Analysis
The relative residence time (rRT) for Patient B is shown in Figure 8 Following simulated mitral valve repair, aortic valve opening improved washout within the aorta Combined MV&AV repair produced the largest and most consistent reductions across compartments, particularly in the LA (up to ∼37%) and LV (up to ∼11%), indicating synergistic improvements in washout when both valves are addressed
Conclusion
Correcting mitral and aortic insufficiency could not only improves left heart hemodynamics and washout efficiency but may also mitigate progressive right ventricular dysfunction in the LVAD population These findings highlight that correcting regurgitant lesions in LVAD patients may prevent progressive RV–PA uncoupling and improve RV loading conditions The patient-specific nature of simulated repair outcomes is evident across the cohort The small sample size (N = 5) limits the ability to draw definitive statistical conclusions
Limitations
The computational framework does not capture dynamic or long-term cardiovascular adaptation Valve function was modeled using simplified planar surfaces with prescribed effective orifice areas rather than fully resolved leaflet anatomy The model does not account for perioperative factors that may influence actual valve repair outcomes, such as surgical technique, annuloplasty ring sizing, residual leaflet tethering, or acute changes in ventricular loading following surgery
Ethics
All data collection and analysis were conducted under IRB-approved protocol HUM00196629 (approved April 2021)
Acknowledgments
This work was supported by the National Institute of Health National Heart, Lung, & Blood Institute [No.
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Improved AI system can perform patient-specific hemodynamic simulations to evaluate valve repair effects in LVAD hearts by integrating dynamic CT imaging, echocardiography, and device parameters into a coupled 3D-0D simulation pipeline.
-
The improved system can simulate how mitral valve repair improves aortic valve opening under LVAD support and how simulated aortic and mitral valve repair improves right ventricle loading conditions.
-
The system can conduct residence time analysis to evaluate blood stasis within the left heart and aorta, revealing that
simulated mitral valve repair restored aortic valve opening during the cardiac cycle, which improved sinus washout.
-
The AI can quantify changes in cardiac output and pulmonary pressures resulting from simulated repairs, such as showing that
simulated mitral valve repair increased by 21.9±16.42% (p = 0.02) and PAPi increased by 24.5±23.55% (p = 0.04).
-
The system can predict the hemodynamic benefits of combined repairs, demonstrating that
Combined in silico mitral and aortic valve repair lowered the LV diastolic pressure more than isolated MV or AV in silico repair.
-
The improved system can map regions of high flow recirculation and quantify relative residence time (rRT) across the left heart and aorta to assess blood stasis risk, showing that
simulated mitral valve repair restored aortic valve opening during the cardiac cycle, which improved sinus washout.
Abstract
Valve dysfunction is a major threat to long-term success in left ventricle assist device (LVAD) therapy, with direct implications for right heart performance. In this study, we apply a patient-specific, image-based computational modeling framework to evaluate the hemodynamic impact of simulated mitral and aortic valve repair in five LVAD-supported patients. Each patient was modeled under four conditions: (patient-specific LVAD-supported state), simulated mitral valve (MV) repair, simulated aortic valve (AV) repair, and simulated combined MV&AV repair. Because validation data for valve repair were unavailable, the simulated repair scenarios are exploratory in silico interventions based on clinically validated patient-specific models. The models integrate dynamic CT imaging, echocardiography, catheterization data, and device-specific LVAD parameters into a coupled 3D-0D simulation pipeline. Valve dynamics are governed by transvalvular pressure and flow, allowing physiological modeling of regurgitant lesions and surgical repair. The right ventricular (RV) was assessed using a combination of model-derived metrics, including right ventricular ejection fraction (RVEF), pulmonary artery pulsatility index (PAPi), and RV-PA coupling. In addition, we performed blood residence time (RT) analysis to evaluate blood stasis within the left heart and aorta. The simulations suggest that valve repair improved cardiac output, reduced pulmonary congestion, and enhanced right ventricular loading conditions. Notably, mitral valve repair restored aortic valve opening during the cardiac cycle, which improved sinus washout and reduced blood residence time within the aortic root-factors associated with lower thrombotic risk. Overall, these findings suggest a potential role for valve repair in LVAD-supported hearts, though larger, validated patient cohorts are needed to confirm these results.
Related papers
- BART Online Open-Source Sequence Toolbox for Computational MRI
- Constrained estimation of rotational invariants of the cumulant expansion (RICE) for rapid tensor-valued diffusion MRI
- Improved Myelin Water Fraction Estimation and Uncertainty Mapping in Multicomponent T2 Relaxometry via NNLS Bootstrap Aggregation
- Radiation damage to normal mammalian tissue in vivo with laser-driven protons at ultra-high instantaneous dose rate
- Fast ungated five-dimensional cardiac MRI on a 1.5 T MR-linac for MRI-guided radiotherapy
- BEAM3R: Beam's-eye-view architecture with Mamba-3 for implicit dose reconstruction