A Multi-scale Linear-time Encoder for Whole-Slide Image Analysis

2026-02-02T23:47:07Z

We introduce Multi-scale Adaptive Recurrent Biomedical Linear-time Encoder (MARBLE), the first \textit{purely Mamba-based} multi-state multiple instance learning (MIL) framework for whole-slide image (WSI) analysis. MARBLE processes multiple magnification levels in parallel and integrates coarse-to-fine reasoning within a linear-time state-space model, efficiently capturing cross-scale dependencies with minimal parameter overhead. WSI analysis remains challenging due to gigapixel resolutions and hierarchical magnifications, while existing MIL methods typically operate at a single scale and transformer-based approaches suffer from quadratic attention costs. By coupling parallel multi-scale processing with linear-time sequence modeling, MARBLE provides a scalable and modular alternative to attention-based architectures. Experiments on five public datasets show improvements of up to \textbf{6.9\%} in AUC, \textbf{20.3\%} in accuracy, and \textbf{2.3\%} in C-index, establishing MARBLE as an efficient and generalizable framework for multi-scale WSI analysis.

Physiology-Informed Generative Multi-Task Network for Contrast-Free CT Perfusion

2026-01-26T17:55:25Z

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.

Quantitative cancer-immunity cycle modeling to optimize bevacizumab and atezolizumab combination therapy for advanced renal cell carcinoma

2026-01-25T03:13:06Z

The incidence of advanced renal cell carcinoma(RCC) has been rising, presenting significant challenges due to the limited efficacy and severe side effects of traditional radiotherapy and chemotherapy. While combination immunotherapies show promise, optimizing treatment strategies remains difficult due to individual heterogeneity. To address this, we developed a Quantitative Cancer-Immunity Cycle (QCIC) model that integrates ordinary differential equations with stochastic modelling to quantitatively characterize and predict tumor evolution in patients with advanced RCC. By systematically integrating quantitative systems pharmacology principles with biological mechanistic knowledge, we constructed a virtual patient cohort and calibrated the model parameters using clinical immunohistochemistry data to ensure biological validity. To enhance predictive performance, we coupled the model with pharmacokinetic equations and defined the Tumor Response Index (TRI) as a quantitative metric of efficacy. Systematic analysis of the QCIC model allowed us to determine an optimal treatment regimen for the combination of bevacizumab and atezolizumab and identify tumor biomarkers with clinical predictive value. This study provides a theoretical framework and methodological support for precision medicine in the treatment of advanced RCC.

A phenotype-structured mathematical model for the influence of hypoxia on oncolytic virotherapy

2026-01-22T17:16:08Z

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.

Transfer Learning from One Cancer to Another via Deep Learning Domain Adaptation

2026-01-21T05:50:34Z

Supervised deep learning models often achieve excellent performance within their training distribution but struggle to generalize beyond it. In cancer histopathology, for example, a convolutional neural network (CNN) may classify cancer severity accurately for cancer types represented in its training data, yet fail on related but unseen types. Although adenocarcinomas from different organs share morphological features that might support limited cross-domain generalization, addressing domain shift directly is necessary for robust performance. Domain adaptation offers a way to transfer knowledge from labeled data in one cancer type to unlabeled data in another, helping mitigate the scarcity of annotated medical images. This work evaluates cross-domain classification performance among lung, colon, breast, and kidney adenocarcinomas. A ResNet50 trained on any single adenocarcinoma achieves over 98% accuracy on its own domain but shows minimal generalization to others. Ensembling multiple supervised models does not resolve this limitation. In contrast, converting the ResNet50 into a domain adversarial neural network (DANN) substantially improves performance on unlabeled target domains. A DANN trained on labeled breast and colon data and adapted to unlabeled lung data reaches 95.56% accuracy. We also examine the impact of stain normalization on domain adaptation. Its effects vary by target domain: for lung, accuracy drops from 95.56% to 66.60%, while for breast and colon targets, stain normalization boosts accuracy from 49.22% to 81.29% and from 78.48% to 83.36%, respectively. Finally, using Integrated Gradients reveals that DANNs consistently attribute importance to biologically meaningful regions such as densely packed nuclei, indicating that the model learns clinically relevant features and can apply them to unlabeled cancer types.

Improving Artifact Robustness for CT Deep Learning Models Without Labeled Artifact Images via Domain Adaptation

2026-01-21T04:17:37Z

If a CT scanner introduces a new artifact not present in the training labels, the model may misclassify the images. Although modern CT scanners include design features which mitigate these artifacts, unanticipated or difficult-to-mitigate artifacts can still appear in practice. The direct solution of labeling images from this new distribution can be costly. As a more accessible alternative, this study evaluates domain adaptation as an approach for training models that maintain classification performance despite new artifacts, even without corresponding labels. We simulate ring artifacts from detector gain error in sinogram space and evaluate domain adversarial neural networks (DANN) against baseline and augmentation-based approaches on the OrganAMNIST abdominal CT dataset. We simulate the absence of labels from an unseen distribution via masking in the loss function and selectively detaching unlabeled instances from the computational graph. Our results demonstrate that baseline models trained only on clean images fail to generalize to images with ring artifacts, and traditional augmentation with other distortion types provides no improvement on unseen artifact domains. In contrast, the DANN approach improves classification accuracy on ring artifact images using only unlabeled artifact data during training, demonstrating the viability of domain adaptation for artifact robustness. The domain-adapted model achieved a classification accuracy of 77.4% on ring artifact test data, 38.7% higher than a baseline model only trained on images with no artifact. These findings provide empirical evidence that domain adaptation can effectively address distribution shift in medical imaging without requiring expensive expert labeling of new artifact distributions, suggesting promise for deployment in clinical settings where novel artifacts may emerge.

Building Digital Twins of Different Human Organs for Personalized Healthcare

2026-01-16T14:15:06Z

Digital twins are virtual replicas of physical entities and are poised to transform personalized medicine through the real-time simulation and prediction of human physiology. Translating this paradigm from engineering to biomedicine requires overcoming profound challenges, including anatomical variability, multi-scale biological processes, and the integration of multi-physics phenomena. This survey systematically reviews methodologies for building digital twins of human organs, structured around a pipeline decoupled into anatomical twinning (capturing patient-specific geometry and structure) and functional twinning (simulating multi-scale physiology from cellular to organ-level function). We categorize approaches both by organ-specific properties and by technical paradigm, with particular emphasis on multi-scale and multi-physics integration. A key focus is the role of artificial intelligence (AI), especially physics-informed AI, in enhancing model fidelity, scalability, and personalization. Furthermore, we discuss the critical challenges of clinical validation and translational pathways. This study not only charts a roadmap for overcoming current bottlenecks in single-organ twins but also outlines the promising, albeit ambitious, future of interconnected multi-organ digital twins for whole-body precision healthcare.

Constitutive parameter inference using physics-based data-driven modeling in full volume datasets of intact and torn rotator cuff tendons

2026-01-14T17:51:44Z

In this work, we characterized the material properties of an animal model of the rotator cuff tendon using full volume datasets of both its intact and injured states by capturing internal strain behavior throughout the tendon. Our experimental setup, involving tension along the fiber direction, activated volumetric, tensile, and shear mechanisms due to the tendon's complex geometry. We implemented an approach to model inference that we refer to as variational system identification (VSI) to solve the weak form of the stress equilibrium equation using these full volume displacements. Three constitutive models were used for parameter inference: a neo-Hookean model, a modified Holzapfel-Gasser-Ogden (HGO) model with higher-order terms in the first and second invariants, and a reduced polynomial model consisting of terms based on the first, second, and fiber-related invariants. Inferred parameters were further refined using an adjoint-based partial differential equation (PDE)-constrained optimization framework. Our results show that the modified HGO model captures the tendon's deformation mechanisms with reasonable accuracy, while the neo-Hookean model fails to reproduce key internal features, particularly the shear behavior in the injured tendon. Surprisingly, the simplified polynomial model performed comparably to the modified HGO formulation using only three terms. These findings suggest that while current constitutive models do not fully replicate the complex internal mechanics of the tendon, they are capable of capturing key trends in both intact and damaged tissue, using a homogeneous modeling approach. Continued model development is needed to bridge this gap and enable clinical-grade, predictive simulations of tendon injury and repair.

Unsupervised full-field Bayesian inference of orthotropic hyperelasticity from a single biaxial test: a myocardial case study

2026-01-14T15:05:54Z

Cardiac muscle tissue exhibits highly non-linear hyperelastic and orthotropic material behavior during passive deformation. Traditional constitutive identification protocols therefore combine multiple loading modes and typically require multiple specimens and substantial handling. In soft living tissues, such protocols are challenged by inter- and intra-sample variability and by manipulation-induced alterations of mechanical response, which can bias inverse calibration. In this work we exploit spatially heterogeneous full-field kinematics as an information-rich alternative to multimodal testing. We adapt EUCLID, an unsupervised method for the automated discovery of constitutive models, towards Bayesian parameter inference for highly nonlinear, orthotropic constitutive models. Using synthetic myocardial tissue slabs, we demonstrate that a single heterogeneous biaxial experiment, combined with sparse reaction-force measurements, enables robust recovery of Holzapfel-Ogden parameters with quantified uncertainty, across multiple noise levels. The inferred responses agree closely with ground-truth simulations and yield credible intervals that reflect the impact of measurement noise on orthotropic material model inference. Our work supports single-shot, uncertainty-aware characterization of nonlinear orthotropic material models from a single biaxial test, reducing sample demand and experimental manipulation.

Simulations Predict Improved Valve Performance Without Direct Leaflet Intervention After Neonatal Truncus Arteriosus Repair

2026-01-13T19:10:40Z

Truncus arteriosus (TA) is a rare and severe congenital heart disease. Quadricuspid valve morphology occurs in 25% of all TA patients and is linked to regurgitation and increased risk of re-operation. It remains unclear how hemodynamic changes after TA repair alter valve performance. This study simulated pre- and postoperative conditions in a neonatal TA patient to investigate valve performance without direct intervention. We hypothesize that valve performance before and after truncal repair can be predicted in-silico, matching in-vivo imaging and identifying mechanisms how hemodynamic changes after repair will reduce valve regurgitation without direct intervention. Pre- and postoperative CT images of a neonatal patient with quadricuspid valve were segmented. Free edge length and geometric height from the patient's echocardiogram were used to model the valve. For the preoperative condition, ventricular pressures were set equal modeling an unrestricted ventricular septal defect. Systemic and pulmonary resistances were tuned based on the patient's Qp:Qs ratio. For the postoperative condition, boundary conditions were modified to mimic patient-specific hemodynamics after TA repair. The preoperative simulation confirmed mild valve regurgitation seen in-vivo. Interaction between asymmetric flow and surrounding vessel resulted in asymmetric opening and closing. Poor central coaptation led to a central regurgitant jet toward the septum. Altered postoperative hemodynamics improved coaptation and eliminated regurgitation, as seen in-vivo. This modeling approach reproduced in-vivo pre- and postoperative valve performance and identified mechanisms improving coaptation after TA repair. TA repair led to elimination of regurgitation due to enhanced central coaptation. Thus, altered postoperative hemodynamic conditions after TA repair may improve valve performance without direct leaflet intervention.

Mechanics of axis formation in $\textit{Hydra}$

2026-01-08T18:45:25Z

The emergence of a body axis is a fundamental step in the development of multicellular organisms. In simple systems such as $\textit{Hydra}$, growing evidence suggests that mechanical forces generated by collective cellular activity play a central role in this process. Here, we explore a physical mechanism for axis formation based on the coupling between active stresses and tissue elasticity. We analyse the elastic deformation induced by activity-generated stresses and show that, owing to the spherical topology of the tissue, forces globally condense toward configurations in which both elastic strain and nematic defect localise at opposite poles. These mechanically selected states define either a polar or apolar head-food axis. To characterize the condensed regime, we introduce a compact parametrization of of the active force and flux distributions, enabling analytical predictions and direct comparison with experiments. Using this framework, we calculate experimentally relevant observables, including areal strain, lateral pressure, and normal displacements during muscular contraction, as well as the detailed structure of topological defect complexes in head and foot regions. Together, our results identify a mechanical route by which active tissues can spontaneously break symmetry at the organismal scale, suggesting a general physical principle underlying body-axis specification during morphogenesis.

Vertical Ground Reaction Forces Waveform Flattening during Gait in Women with Knee Osteoarthritis

2026-01-07T20:09:19Z

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

The mechanics of the $\textit{Less In More Out}$ artificial heart: modeling fabric-based soft robotic devices

2026-01-07T09:43:10Z

Recently, the Less In More Out device, a fluidically actuated soft total artificial heart was proposed. This device uses arrays of pouch motors to achieve a positive fluidic lever when pneumatically actuated against physiological hemodynamic conditions. Extensive experimental characterization demonstrated its potential; however, experiments alone cannot resolve the internal mechanical fields that govern device durability and performance. Here, we develop a computational framework to investigate intrinsic device mechanics, such as stress concentrations, strain paths, and fatigue life, and to explore targeted design modifications that improve durability and efficiency. We show that our model reproduces the nonlinear deformation and pressure-volume relationships measured experimentally under varying hemodynamic conditions. Across designs, devices with fewer pouches deliver higher stroke volumes but exhibit up to 50% higher peak von Mises stresses, which reduces their fatigue life. Our simulations further identify heat-sealed seams and buckling regions as durability-limiting features. As a proof of concept, we vary the valve support aspect ratio and relative endocardial-epicardial pouch fabric compliance, reducing the peak von Mises stress by ~10% while maintaining identical physiological outputs and improving mechanical efficiency. Overall, our framework enables detailed evaluation of stress hotspots, buckling, and fatigue life, and offers a foundation for optimizing artificial hearts and other fluidically actuated fabric-based soft robotic devices.

Quantifying Local Strain Field and Deformation in Active Contraction of Bladder Using a Pretrained Transformer Model: A Speckle-Free Approach

2026-01-04T00:52:27Z

Accurate quantification of local strain fields during bladder contraction is essential for understanding the biomechanics of bladder micturition, in both health and disease. Conventional digital image correlation (DIC) methods have been successfully applied to various biological tissues; however, this approach requires artificial speckling, which can alter both passive and active properties of the tissue. In this study, we introduce a speckle-free framework for quantifying local strain fields using a state-of-the-art, zero-shot transformer model, CoTracker3. We utilized a custom-designed, portable isotonic biaxial apparatus compatible with multiphoton microscopy (MPM) to demonstrate this approach, successfully tracking natural bladder lumen textures without artificial markers. Benchmark tests validated the method's high pixel accuracy and low strain errors. Our framework effectively captured heterogeneous deformation patterns, despite complex folding and buckling, which conventional DIC often fails to track. Application to in vitro active bladder contractions in four rat specimens (n=4) revealed statistically significant anisotropy (p<0.01), with higher contraction longitudinally compared to circumferentially. Multiphoton microscopy further illustrated and confirmed heterogeneous morphological changes, such as large fold formation during active contraction. This non-invasive approach eliminates speckle-induced artifacts, enabling more physiologically relevant measurements, and has broad applicability for material testing of other biological and engineered systems.

Model-Assisted Causal Inference for the Treatment Effect on Recurrent Events in the Presence of Terminal Events

2026-01-03T17:40:33Z

This paper is motivated by evaluating the benefits of patients receiving mechanical circulatory support (MCS) devices in end-stage heart failure management inference, in which hypothesis testing for a treatment effect on the risk of recurrent events is challenged in the presence of terminal events. Existing methods based on cumulative frequency unreasonably disadvantage longer survivors as they tend to experience more recurrent events. The While-Alive-based (WA) test has provided a solution to address this survival-length-bias problem, and it performs well when the recurrent event rate holds constant over time. However, if such a constant-rate assumption is violated, the WA test can exhibit an inflated type I error and inaccurate estimation of treatment effects. To fill this methodological gap, we propose a Proportional Rate Marginal Structural Model-assisted Test (PR-MSMaT) in the causal inference framework of separable treatment effects for recurrent and terminal events. Using the simulation study, we demonstrate that our PR-MSMaT can properly control type I error while gaining power comparable to the WA test under time-varying recurrent event rates. We employ PR-MSMaT to compare different MCS devices with the postoperative risk of gastrointestinal bleeding among patients enrolled in the Interagency Registry of Mechanically Assisted Circulatory Support program.