arXiv subjects

Mia Bonini

Publications and source records attributed to Mia Bonini.

7 recordsLinked to original sources

arXiv2026

Spectrally optimized implicit neural representations mitigate acquisition-related image quality trade-offs in intracranial 4D Flow MRI

4D Flow MRI promises opportunities for non-invasive mapping of neurovascular hemodynamics, but its clinical utility faces trade-offs between resolution, acquisition time, and noise. Deep learning-based super-resolution offers a post-scan solution, but supervised approaches require paired or domain-specific data, while patient-specific physics-informed models can be computationally demanding. This creates a need for unsupervised reconstruction models that can adapt to individual acquisitions without paired high-resolution training data, while remaining robust across anatomy, disease severity, and acquisition quality. To address this, we introduce a spectrally tuned implicit neural representation (INR) framework for intracranial 4D Flow MRI, representing the velocity field as a continuous coordinate-based neural field using WIRE activation functions with tunable spectral bias. Data-driven and physics-informed variants were evaluated in nine cerebrovascular cases spanning healthy controls and moderate-to-severe intracranial atherosclerotic disease, using CFD-derived high-resolution reference fields, MR-realistic synthetic low-resolution inputs, and direct fitting to acquired in-vivo data. Cohort-optimal spectral-bias settings generalized across subjects with only small increases in relative error compared with subject-specific optima, reducing the need for case-specific hyperparameter tuning. Physics-informed regularization provided modest additional accuracy but required substantially longer optimization, whereas the data-driven INR achieved near-comparable reconstruction performance within minutes. The framework also remained robust to acquisition-realistic noise and moderate VENC-dependent aliasing. These findings establish spectrally optimized INRs as a flexible unsupervised approach for mitigating acquisition-related trade-offs in intracranial 4D Flow MRI.

eess.IV↗

arXiv2026

Comprehensive Analysis of Relative Pressure Estimation in Type B Aortic Dissection Utilizing 4D-Flow MRI

Magnetic resonance imaging (MRI) can estimate three-dimensional (3D) time-resolved blood velocity fields using 4D-flow MRI, providing rich relative pressure field information. Clinical alternatives---catheterization and Doppler echocardiography---provide one-dimensional pressure drops. While the accuracy of one-dimensional pressure drops from 4D-flow has been explored, the accuracy of 3D relative pressure field estimates requires further evaluation. This work analyzes three state-of-the-art relative pressure estimators: virtual Work-Energy Relative Pressure (vWERP), the Pressure Poisson Estimator (PPE), and the Stokes Estimator (STE). Spatiotemporal characteristics and noise sensitivity were determined in silico via comparison with CFD-derived relative pressure fields. Estimators were then validated using a type B aortic dissection (TBAD) flow phantom with varying tear geometry and twelve catheter pressure measurements. Finally, each estimator was evaluated across eight patient cases in a clinical feasibility analysis. Estimators were cross-compared and evaluated for coherence since no reference catheter pressure measurements were available. PPE outperformed STE in the in silico relative pressure analysis, while both outperformed vWERP in plane-to-plane assessment. High velocity gradients and low spatial resolution contributed most significantly to local variations in 3D relative pressure field errors. Low temporal resolution led to systematic underestimation of highly transient peak pressure events. In the flow phantom analysis, vWERP was the most accurate method, followed by STE and then PPE. Each relative pressure estimator was strongly correlated with ground truth relative pressure values, despite the tendency to underestimate peak relative pressure. Patient case results demonstrated that each relative pressure estimator could be feasibly integrated into a clinical workflow.

physics.flu-dyn↗

arXiv2026

Towards Modeling the Hemodynamic Impact of Mitral and Aortic Valve Repair in Patients with Left Ventricular Assist Devices

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.

physics.med-ph↗

arXiv2025

In Silico Evaluation of Cardiac Tissue-Engineered Patch Interventions

Myocardial infarction significantly degrades heart function, and current treatments can bring forth serious cost and complications including blood clots and infections. To improve the current state of treatment, researchers are developing tissue patches from induced-pluripotent stem cells that can be incorporated into the heart, improving organ function after a myocardial infarction. These tissue patches include surface patches, attached to the epicardium of the heart, and thick transmural patches that replace the infarcted region. However, little is known about the impact of cardiac tissue patches on pump function in a patient's heart. In addition, it is not clear what patch structural properties - such as active stress generation, muscle fiber alignment, or material stiffness - may best augment existing heart tissue. Computational modeling can be used to examine different implementations and patch properties, illuminating the mechanical impact of cardiac tissue patches in the beating heart. In this work, we computationally implement different cardiac tissue patches to understand benefits of particular patch types and properties. We find that in transmural cardiac tissue patches, both activation and fiber alignment improve function. A transmural patch generating 10% of healthy active stress can increase stroke volume by 18%, and higher generated active stress in a circumferential muscle fiber orientation can recover stroke volume by over 50%. Furthermore, we find that surface cardiac tissue patches can enhance heart function slightly despite limiting diastolic filling, especially when fibrotic thinning has occurred. These conclusions identify broad design goals for the engineering of cardiac tissue patches to improve heart function after a myocardial infarction.

physics.med-ph↗

arXiv2025

Deep learning for temporal super-resolution 4D Flow MRI

4D Flow Magnetic Resonance Imaging (4D Flow MRI) is a non-invasive technique for volumetric, time-resolved blood flow quantification. However, apparent trade-offs between acquisition time, image noise, and resolution limit clinical applicability. In particular, in regions of highly transient flow, coarse temporal resolution can hinder accurate capture of physiologically relevant flow variations. To overcome these issues, post-processing techniques using deep learning have shown promising results to enhance resolution post-scan using so-called super-resolution networks. However, while super-resolution has been focusing on spatial upsampling, temporal super-resolution remains largely unexplored. The aim of this study was therefore to implement and evaluate a residual network for temporal super-resolution 4D Flow MRI. To achieve this, an existing spatial network (4DFlowNet) was re-designed for temporal upsampling, adapting input dimensions, and optimizing internal layer structures. Training and testing were performed using synthetic 4D Flow MRI data originating from patient-specific in-silico models, as well as using in-vivo datasets. Overall, excellent performance was achieved with input velocities effectively denoised and temporally upsampled, with a mean absolute error (MAE) of 1.0 cm/s in an unseen in-silico setting, outperforming deterministic alternatives (linear interpolation MAE = 2.3 cm/s, sinc interpolation MAE = 2.6 cm/s). Further, the network synthesized high-resolution temporal information from unseen low-resolution in-vivo data, with strong correlation observed at peak flow frames. As such, our results highlight the potential of utilizing data-driven neural networks for temporal super-resolution 4D Flow MRI, enabling high-frame-rate flow quantification without extending acquisition times beyond clinically acceptable limits.

cs.LG↗

arXiv2024

Effective Block Preconditioners for Fluid Dynamics Coupled to Reduced Models of a Non-Local Nature

Modeling cardiovascular blood flow is central to many applications in biomedical engineering. To accommodate the complexity of the cardiovascular system, in terms of boundary conditions and surrounding vascular tissue, computational fluid dynamics (CFD) often are coupled to reduced circuit and/or solid mechanics models. These allow for realistic simulations of hemodynamics in the heart or the aorta, but come at additional computational cost and complexity. In this contribution, we design a novel block preconditioner for the solution of the stabilized Navier-Stokes equations coupled to reduced-order models of a non-local nature. These models encompass lumped-parameter systems that impose flux-dependent boundary tractions, and Galerkin reduced-order models that can be used to account for outlying mechanical structures. Here we propose a 3x3 preconditioner derived from the block factorization and approximation to the Schur complement(s). The solver performance is demonstrated for a series of examples with increasing complexity, culminating in a reduced FSI simulation in a patient-specific contracting left heart model. For all test cases, we show that our proposed approach is superior to other frequently presented 2x2 schemes that merge stiffness contributions from reduced models into the fluid Jacobian or consolidate some variables for the purpose of efficiency - with an up to six times shorter overall computing time and/or only half as many linear iterations.

physics.flu-dyn↗

arXiv2023

Generalized super-resolution 4D Flow MRI $\unicode{x2013}$ using ensemble learning to extend across the cardiovascular system

4D Flow Magnetic Resonance Imaging (4D Flow MRI) is a non-invasive measurement technique capable of quantifying blood flow across the cardiovascular system. While practical use is limited by spatial resolution and image noise, incorporation of trained super-resolution (SR) networks has potential to enhance image quality post-scan. However, these efforts have predominantly been restricted to narrowly defined cardiovascular domains, with limited exploration of how SR performance extends across the cardiovascular system; a task aggravated by contrasting hemodynamic conditions apparent across the cardiovasculature. The aim of our study was to explore the generalizability of SR 4D Flow MRI using a combination of heterogeneous training sets and dedicated ensemble learning. With synthetic training data generated across three disparate domains (cardiac, aortic, cerebrovascular), varying convolutional base and ensemble learners were evaluated as a function of domain and architecture, quantifying performance on both in-silico and acquired in-vivo data from the same three domains. Results show that both bagging and stacking ensembling enhance SR performance across domains, accurately predicting high-resolution velocities from low-resolution input data in-silico. Likewise, optimized networks successfully recover native resolution velocities from downsampled in-vivo data, as well as show qualitative potential in generating denoised SR-images from clinical level input data. In conclusion, our work presents a viable approach for generalized SR 4D Flow MRI, with ensemble learning extending utility across various clinical areas of interest.

eess.IV↗