31 papers
Weakly nonlinear analysis of a reaction-diffusion model for demyelinating lesions in Multiple Sclerosis
Multiple Sclerosis is a chronic autoimmune disorder characterized by the degradation of the myelin sheath in the central nervous system, leading to neurological impairments. In this work, we analyze a reaction-diffusion model derived from kinetic theory to study the formation of demyelinating lesions. We perform a Turing instability analysis and a weakly nonlinear analysis to investigate different spatial patterns that may emerge. Our study examines how key parameters, including the squeezing probability of immune cells and the chemotactic response, impact pattern formation. Numerical simulations confirm the analytical results, revealing the emergence of distinct spatial structures.
2603.06628Feb 2026
View
Spatio-temporal air flow properties in a 3D personalised model of the human lung
We propose a multi-scale lung model to investigate spatio-temporal distributions of ventilation variables. Lung envelope and large airway geometries are derived from CT scans; smaller airways are generated using a physiologically consistent algorithm. Tissue mechanics is modeled using nonlinear elasticity under small deformations, coupled with local air pressure from fluid dynamics within the bronchial tree. Airflow accounts for inertia and static airway compliance. Simulations employ finite elements. Using this model, we explore spatio-temporal airflows and shear stresses distributions.
2602.17265Feb 2026
View
A dialog between cell adhesion and topology at the core of morphogenesis
During the development of an organism, cells must coordinate and organize to generate the correct shape, structure, and spatial patterns of tissues and organs, a process known as morphogenesis. The morphogenesis of embryonic tissues is supported by multiple processes that induce the precise physical deformations required for tissues to ultimately form organs with complex geometries. Among the most active players shaping the morphogenetic path are fine-tuned changes in cell adhesion. In this paper, we show that a local, pair-wise property defined at the cell-cell contact level has important global consequences for embryonic tissue topology, being determinant in defining both the geometric and material properties of early embryo tissues.
2602.09867Feb 2026
View
Mathematical Modeling of Lesion Pattern Formation in Dendritic Keratitis
Dendritic keratitis is a form of eye infection caused by herpes simplex virus (HSV). The virus spreads via direct cell-to-cell infection among corneal epithelial cells. This leads to the formation of dendritic lesions characterized by terminal bulbs at their tips. Under immunosuppression, the condition may progress to geographic keratitis, which is a map-shaped lesion with dendritic tails. The mechanism of this pattern formation remains to be elucidated. In this study, we propose a mathematical model to elucidate the mechanisms of lesion pattern formation in dendritic keratitis. Our model shows that increased production of infection-suppressive cytokines induces dendritic patterns with terminal bulbs, whereas reduced cytokine levels lead to geographic patterns. Furthermore, altering the spatial distribution of cytokine production can reproduce dendritic tails. By including external cytokine secretion, we could reproduce tapered lesions observed in non-HSV keratitis. By clarifying the mechanisms behind terminal bulb formation and reproducing atypical lesion morphologies, our findings enhance the understanding of herpetic keratitis and highlight the utility of mathematical modeling in ophthalmology.
2602.02916Feb 2026
View
Fast generation of 3D flow obstacles from parametric surface models: application to cardiac valves
Due to the computationally demanding nature of fluid-structure interaction simulations, heart valve simulation is a complex task. A simpler alternative is to model the valve as a resistive flow obstacle that can be updated dynamically without altering the mesh, but this approach can also become computationally expensive for large meshes. In this work, we present a fast method for computing the resistive flow obstacle of a heart valve. The method is based on a parametric surface model of the valve, which is defined by a set of curves. The curves are adaptively sampled to create a polyline representation, which is then used to generate the surface. The surface is represented as a set of points, allowing for efficient distance calculations to determine whether mesh nodes belong to the valve surface. We introduce three algorithms for computing these distances: minimization, sampling, and triangulation. Additionally, we implement two mesh traversal strategies: exhaustive node iteration and recursive neighbor search. The latter significantly reduces the number of distance calculations by only considering neighboring nodes. Our pipeline is demonstrated on both a previously reported aortic valve model and a newly proposed mitral valve model, highlighting its flexibility and efficiency for rapid valve shape updates in computational simulations.
2512.09945Dec 2025
View
Vertical Ground Reaction Forces Waveform Flattening during Gait in Women with Knee Osteoarthritis
Background. Knee Osteoarthritis (OA) is a common chronic joint condition, and its prevalence increases with age. This study aims to examine whether flattened vertical ground reaction force (vGRF) waveforms and reduced knee range of motion (RoM) occur together during gait as compensatory strategies to maintain gait speed. Methods. Twelve women with knee OA and twelve healthy women of the same age completed the Western Ontario and McMaster University Index (WOMAC) to assess self-reported pain, stiffness, and function. The groups were divided into two groups: OA vs. control 2 limbs or left and right in the Control group. A mixed-design ANOVA was used to examine differences in vertical ground response forces (VGRFs) peaks, minimum VGRF, anterior-posterior weight acceptance (ADWA) and propulsive force (ADPO), knee RoM, and gait speeds. Results. In the OA group, the mean Peak 1 vGFR was 1.109 (SD = 0.05) for the right leg (p 0.05), while the mean min vGFFR was 0.87 (SD=0.04) for the left leg. The OA leg exhibited a mean ADWA of 0.23 0.04 kg/BW, which was significantly lower than the control group's right leg (0.28 0.09 kg/bw, p0.05). No group differences in gait velocity were detected. Conclusions. We interpret the flattening of the vFGFR waveform and the reduction in knee RoM as components of an adaptive, yet potentially maladaptive, motor strategy
2510.16942Oct 2025
View
Unsupervised full-field Bayesian inference of orthotropic hyperelasticity from a single biaxial test: a myocardial case study
Fully capturing this behavior in traditional homogenized tissue testing requires the excitation of multiple deformation modes, i.e. combined triaxial shear tests and biaxial stretch tests. Inherently, such multimodal experimental protocols necessitate multiple tissue samples and extensive sample manipulations. Intrinsic inter-sample variability and manipulation-induced tissue damage might have an adverse effect on the inversely identified tissue behavior. In this work, we aim to overcome this gap by focusing our attention to the use of heterogeneous deformation profiles in a parameter estimation problem. More specifically, we adapt EUCLID, an unsupervised method for the automated discovery of constitutive models, towards the purpose of parameter identification for highly nonlinear, orthotropic constitutive models using a Bayesian inference approach and three-dimensional continuum elements. We showcase its strength to quantitatively infer, with varying noise levels, the material model parameters of synthetic myocardial tissue slabs from a single heterogeneous biaxial stretch test. This method shows good agreement with the ground-truth simulations and with corresponding credibility intervals. Our work highlights the potential for characterizing highly nonlinear and orthotropic material models from a single biaxial stretch test with uncertainty quantification.
2510.09498Oct 2025
View
2509.16328The Role of High-Performance GPU Resources in Large Language Model Based Radiology Imaging Diagnosis
Large-language models (LLMs) are rapidly being applied to radiology, enabling automated image interpretation and report generation tasks. Their deployment in clinical practice requires both high diagnostic accuracy and low inference latency, which in turn demands powerful hardware. High-performance graphical processing units (GPUs) provide the necessary compute and memory throughput to run large LLMs on imaging data. We review modern GPU architectures (e.g. NVIDIA A100/H100, AMD Instinct MI250X/MI300) and key performance metrics of floating-point throughput, memory bandwidth, VRAM capacity. We show how these hardware capabilities affect radiology tasks: for example, generating reports or detecting findings on CheXpert and MIMIC-CXR images is computationally intensive and benefits from GPU parallelism and tensor-core acceleration. Empirical studies indicate that using appropriate GPU resources can reduce inference time and improve throughput. We discuss practical challenges including privacy, deployment, cost, power and optimization strategies: mixed-precision, quantization, compression, and multi-GPU scaling. Finally, we anticipate that next-generation features (8-bit tensor cores, enhanced interconnect) will further enable on-premise and federated radiology AI. Advancing GPU infrastructure is essential for safe, efficient LLM-based radiology diagnostics.
2509.16328Sep 2025
View
The mechanics of $\textit{Less In More Out}$: modeling fabric-based soft robotic hearts
Fabric-based soft robots combine high load-carrying capacity, efficiency, and low weight with the ability to bend, twist, contract, or extend with ease, making them promising candidates for biomedical applications such as soft total artificial hearts. While recent experiments have demonstrated their potential, predictive numerical models are urgently needed to study their complex mechanics, guide design optimization and improve their reliability. We develop a computational model of the Less In More Out device, a fluidically actuated soft total artificial heart constructed from heat-sealed layers of woven fabric. Our model reproduces the nonlinear deformation, strain fields, and pressure-volume relationships measured in quasi-static experiments. Devices with fewer pouches deliver higher stroke volumes but exhibit up to 50% higher peak von Mises stresses. Fatigue analysis using a strain-life approach identifies heat-sealed seams and buckling regions as durability-limiting features. Our framework enables detailed evaluation of stress concentrations, buckling, and fatigue life, providing mechanistic insights that are difficult to obtain experimentally. It also offers a foundation for the optimization of artificial hearts and other fluid actuated fabric-based soft robotic systems.
2510.14984Sep 2025
View
Integrated Open-Source Framework for Simulation of Transcatheter Pulmonary Valves in Native Right Ventricular Outflow Tracts
Background - Pulmonary insufficiency is a consequence of transannular patch repair in Tetralogy of Fallot (ToF), leading to late morbidity and mortality. Transcatheter native outflow tract pulmonary valve replacement (TPVR) has become common, but assessment of patient candidacy and selection of the optimal device remains challenging. We demonstrate an integrated open-source workflow for simulation of TPVR in image-derived models to inform device selection. Methods - Machine learning-based segmentation of CT scans was implemented to define the right ventricular outflow tract (RVOT). A custom workflow for device positioning and pre-compression was implemented in SlicerHeart. Resulting geometries were exported to FEBio for simulation. Visualization of results and quantification were performed using custom metrics implemented in SlicerHeart and FEBio. Results - RVOT model creation and device placement could be completed in under 1 minute. Virtual device placement using FE simulations visually mimicked actual device placement and allowed quantification of vessel strain, stress, and contact area. Regions of higher strain and stress were observed at the proximal and distal end locations of the TPVs where the devices impinge the RVOT wall. No other consistent trends were observed across simulations. The observed variability in mechanical metrics across RVOTS, stents, and locations in the RVOT highlights that no single device performs optimally in all anatomies, thereby reinforcing the need for simulation-based patient-specific assessment. Conclusions - This study demonstrates the feasibility of a novel open-source workflow for the rapid simulation of TPVR which with further refinement may inform assessment of patient candidacy and optimal device selection.
2507.063371Jul 2025
View
Physiology-Informed Generative Multi-Task Network for Contrast-Free CT Perfusion
Perfusion imaging is extensively utilized to assess hemodynamic status and tissue perfusion in various organs. Computed tomography perfusion (CTP) imaging plays a key role in the early assessment and planning of stroke treatment. While CTP provides essential perfusion parameters to identify abnormal blood flow in the brain, the use of contrast agents in CTP can lead to allergic reactions and adverse side effects, along with costing USD 4.9 billion worldwide in 2022. To address these challenges, we propose a novel deep learning framework called Multitask Automated Generation of Intermodal CT perfusion maps (MAGIC). This framework combines generative artificial intelligence and physiological information to map non-contrast computed tomography (CT) imaging to multiple contrast-free CTP imaging maps. We demonstrate enhanced image fidelity by incorporating physiological characteristics into the loss terms. Our network was trained and validated using CT image data from patients referred for stroke at UF Health and demonstrated robustness to abnormalities in brain perfusion activity. A double-blinded study was conducted involving seven experienced neuroradiologists and vascular neurologists. This study validated MAGIC's visual quality and diagnostic accuracy showing favorable performance compared to clinical perfusion imaging with intravenous contrast injection. Overall, MAGIC holds great promise in revolutionizing healthcare by offering contrast-free, cost-effective, and rapid perfusion imaging.
2505.22673May 2025
View
RP-SAM2: Refining Point Prompts for Stable Surgical Instrument Segmentation
Accurate surgical instrument segmentation is essential in cataract surgery for tasks such as skill assessment and workflow optimization. However, limited annotated data makes it difficult to develop fully automatic models. Prompt-based methods like SAM2 offer flexibility yet remain highly sensitive to the point prompt placement, often leading to inconsistent segmentations. We address this issue by introducing RP-SAM2, which incorporates a novel shift block and a compound loss function to stabilize point prompts. Our approach reduces annotator reliance on precise point positioning while maintaining robust segmentation capabilities. Experiments on the Cataract1k dataset demonstrate that RP-SAM2 improves segmentation accuracy, with a 2% mDSC gain, a 21.36% reduction in mHD95, and decreased variance across random single-point prompt results compared to SAM2. Additionally, on the CaDIS dataset, pseudo masks generated by RP-SAM2 for fine-tuning SAM2's mask decoder outperformed those generated by SAM2. These results highlight RP-SAM2 as a practical, stable and reliable solution for semi-automatic instrument segmentation in data-constrained medical settings. The code is available at https://github.com/BioMedIA-MBZUAI/RP-SAM2.
2504.071171Mar 2025
View
A mathematical perspective on the paradox that chemotherapy sometimes works backwards
Doctors are well aware that sometimes cancer treatments not only fail, but even work backwards, i.e. they make the treated tumor grow. In this work we present a mathematical perspective on this paradox in the case of chemotherapy, by studying a minimally parameterized mathematical model for the system composed of the tumor and the surrounding vasculature. To this end, we will use a system of two well-established nonlinear ordinary differential equations, which incorporates the cytotoxic (via the Norton-Simon hypothesis) and antiangiogenic effects of chemotherapy. Finally, we provide two theoretical ways to avoid these anomalies.
2503.10670Mar 2025
View
DDCSR: A Novel End-to-End Deep Learning Framework for Cortical Surface Reconstruction from Diffusion MRI
Diffusion MRI (dMRI) plays a crucial role in studying brain white matter connectivity. Cortical surface reconstruction (CSR), including the inner whiter matter (WM) and outer pial surfaces, is one of the key tasks in dMRI analyses such as fiber tractography and multimodal MRI analysis. Existing CSR methods rely on anatomical T1-weighted data and map them into the dMRI space through inter-modality registration. However, due to the low resolution and image distortions of dMRI data, inter-modality registration faces significant challenges. This work proposes a novel end-to-end learning framework, DDCSR, which for the first time enables CSR directly from dMRI data. DDCSR consists of two major components, including: (1) an implicit learning module to predict a voxel-wise intermediate surface representation, and (2) an explicit learning module to predict the 3D mesh surfaces. Compared to several baseline and advanced CSR methods, we show that the proposed DDCSR can largely increase both accuracy and efficiency. Furthermore, we demonstrate a high generalization ability of DDCSR to data from different sources, despite the differences in dMRI acquisitions and populations.
2503.03790Mar 2025
View
Periodontal Bone Loss Analysis via Keypoint Detection With Heuristic Post-Processing
This study proposes a deep learning framework and annotation methodology for the automatic detection of periodontal bone loss landmarks, associated conditions, and staging. 192 periapical radiographs were collected and annotated with a stage agnostic methodology, labelling clinically relevant landmarks regardless of disease presence or extent. We propose a heuristic post-processing module that aligns predicted keypoints to tooth boundaries using an auxiliary instance segmentation model. An evaluation metric, Percentage of Relative Correct Keypoints (PRCK), is proposed to capture keypoint performance in dental imaging domains. Four donor pose estimation models were adapted with fine-tuning for our keypoint problem. Post-processing improved fine-grained localisation, raising average PRCK^{0.05} by +0.028, but reduced coarse performance for PRCK^{0.25} by -0.0523 and PRCK^{0.5} by -0.0345. Orientation estimation shows excellent performance for auxiliary segmentation when filtered with either stage 1 object detection model. Periodontal staging was detected sufficiently, with the best mesial and distal Dice scores of 0.508 and 0.489, while furcation involvement and widened periodontal ligament space tasks remained challenging due to scarce positive samples. Scalability is implied with similar validation and external set performance. The annotation methodology enables stage agnostic training with balanced representation across disease severities for some detection tasks. The PRCK metric provides a domain-specific alternative to generic pose metrics, while the heuristic post-processing module consistently corrected implausible predictions with occasional catastrophic failures. The proposed framework demonstrates the feasibility of clinically interpretable periodontal bone loss assessment, with potential to reduce diagnostic variability and clinician workload.
2503.134771Mar 2025
View
Passive Heart Rate Monitoring During Smartphone Use in Everyday Life
Resting heart rate (RHR) is an important biomarker of cardiovascular health and mortality, but tracking it longitudinally generally requires a wearable device, limiting its availability. We present PHRM, a deep learning system for passive heart rate (HR) and RHR measurements during everyday smartphone use, using facial video-based photoplethysmography. Our system was developed using 225,773 videos from 495 participants and validated on 185,970 videos from 205 participants in laboratory and free-living conditions, representing the largest validation study of its kind. Compared to reference electrocardiogram, PHRM achieved a mean absolute percentage error (MAPE) < 10% for HR measurements across three skin tone groups of light, medium and dark pigmentation; MAPE for each skin tone group was non-inferior versus the others. Daily RHR measured by PHRM had a mean absolute error < 5 bpm compared to a wearable HR tracker, and was associated with known risk factors. These results highlight the potential of smartphones to enable passive and equitable heart health monitoring.
2503.037833Mar 2025
View
A functional exchange shunt in the umbilical cord: the role of coiling in solute and heat transfer
The umbilical cord plays a critical role in delivering nutrients and oxygen from the placenta to the fetus through the umbilical vein, while the two umbilical arteries carry deoxygenated blood with waste products back to the placenta. Although solute exchange in the placenta has been extensively studied, exchange within the cord tissue has not been investigated. Here, we explore the hypothesis that the coiled structure of the umbilical cord could strengthen diffusive coupling between the arteries and the vein, resulting in a functional shunt. We calculate the diffusion of solutes, such as oxygen, and heat in the umbilical cord to quantify how this shunt is affected by vascular configuration within the cord. We demonstrate that the shunt is enhanced by coiling and vessel proximity. Furthermore, our model predicts that typical vascular configurations of the human cord tend to minimise shunting, which could otherwise disrupt thermal regulation of the fetus. We also show that the exchange, amplified by coiling, can provide additional oxygen supply to the cord tissue surrounding the umbilical vessels.
2502.14228Feb 2025
View
Prediction of Lung Metastasis from Hepatocellular Carcinoma using the SEER Database
Hepatocellular carcinoma (HCC) is a leading cause of cancer-related mortality, with lung metastases being the most common site of distant spread and significantly worsening prognosis. Despite the growing availability of clinical and demographic data, predictive models for lung metastasis in HCC remain limited in scope and clinical applicability. In this study, we develop and validate an end-to-end machine learning pipeline using data from the Surveillance, Epidemiology, and End Results (SEER) database. We evaluated three machine learning models (Random Forest, XGBoost, and Logistic Regression) alongside a multilayer perceptron (MLP) neural network. Our models achieved high AUROC values and recall, with the Random Forest and MLP models demonstrating the best overall performance (AUROC = 0.82). However, the low precision across models highlights the challenges of accurately predicting positive cases. To address these limitations, we developed a custom loss function incorporating recall optimization, enabling the MLP model to achieve the highest sensitivity. An ensemble approach further improved overall recall by leveraging the strengths of individual models. Feature importance analysis revealed key predictors such as surgery status, tumor staging, and follow up duration, emphasizing the relevance of clinical interventions and disease progression in metastasis prediction. While this study demonstrates the potential of machine learning for identifying high-risk patients, limitations include reliance on imbalanced datasets, incomplete feature annotations, and the low precision of predictions. Future work should leverage the expanding SEER dataset, improve data imputation techniques, and explore advanced pre-trained models to enhance predictive accuracy and clinical utility.
2501.11720Jan 2025
View
The Mean-Field Survival Model for Stripe Formation in Zebrafish Exhibits Turing Instability
Zebrafish have been used as a model organism in many areas of biology, including the study of pattern formation. The mean-field survival model is a coupled ODE system describing the expected evolution of chromatophores coordinating to form stripes in zebrafish. This paper presents analysis of the model focusing on parameters for the number of cells, length of distant-neighbor interactions, and rates related to birth and death of chromatophores. We derive the conditions on these parameters for a Turing bifurcation to occur and show that the model predicts patterns qualitatively similar to those in nature. In addition to answering questions about this particular model, this paper also serves as a case study for Turing analysis on coupled ODE systems. The qualitative behavior of such coupled ODE models may deviate significantly from continuum limit models. The ability to analyze such systems directly avoids this concern and allows for a more accurate description of the behavior at physically relevant scales.
2411.15293Nov 2024
View
A phenotype-structured mathematical model for the influence of hypoxia on oncolytic virotherapy
The effectiveness of oncolytic virotherapy is significantly affected by several elements of the tumour microenvironment, which reduce the ability of the virus to infect cancer cells. In this work, we focus on the influence of hypoxia on this therapy and develop a novel continuous mathematical model that considers both the spatial and epigenetic heterogeneity of the tumour. We investigate how oxygen gradients within tumours affect the spatial distribution and replication of both the tumour and oncolytic viruses, focusing on regions of severe hypoxia versus normoxic areas. Additionally, we analyse the evolutionary dynamics of tumour cells under hypoxic conditions and their influence on susceptibility to viral infection. Our findings show that the reduced metabolic activity of hypoxic cells may significantly impact the virotherapy effectiveness; the knowledge of the tumour's oxygenation could, therefore, suggest the most suitable type of virus to optimise the outcome. The combination of numerical simulations and theoretical results for the model equilibrium values allows us to elucidate the complex interplay between viruses, tumour evolution and oxygen dynamics, ultimately contributing to developing more effective and personalised cancer treatments.
2411.02413Oct 2024
ViewPage 1 of 2
