Accurate non-invasive localization of right ventricular (RV) arrhythmia origins remains a challenge in electrophysiology. This study investigates the feasibility of using machine learning models based on 12-lead ECG QRS integrals to localize early RV activation sites.

Spinal cord injury (SCI) is a devastating condition with limited therapeutic options owing to poor intrinsic regeneration and the formation of glial scars. Platelet-rich plasma (PRP)-based biomaterials have emerged as promising candidates for neural repair; however, their application in complete SCI models with rigorous multimodal validation has not yet been investigated. This study aimed to extend the current knowledge on SCI-PRP based study, by developing and comprehensively validating a clinically translatable PRP-derived fibrin scaffold for spinal cord regeneration.

Clinical failure rates associated with in-stent restenosis are difficult to predict and manage, particularly at the patient-specific level. Studies have linked biomechanical factors to focal disease development and progression, suggesting that physics-based simulations using finite element (FE) approaches hold potential to mitigate stent failure rates. However, insufficient validation to assess the accuracy of model predictions limit model credibility for clinical translation. Herein, we established a computational framework to validate vascular stent deployment by integrating robust simulation and rigorous experimental approaches.

Neonatal jaundice is a common condition that can lead to serious complications if left untreated. Although phototherapy is the standard treatment, existing devices are often costly, energy-intensive and fail to deliver the required intensity or provide uniform irradiance, particularly in low-resource settings. This study aims to design and optimize a low-cost, energy-efficient phototherapy bed that meets clinical requirements for irradiance intensity and spatial uniformity.

This study aims to evaluate an adapted needs-finding process, termed "Caredesign," which focuses on identifying service delivery and administrative inefficiencies in low- and middle-income countries (LMICs). The research question is whether modifying the traditional Biodesign needs-finding phase can yield actionable insights for healthcare service innovation in resource-constrained settings.

Aortic dissection (AD) is a catastrophic vascular pathology caused by delamination of the vascular wall and the formation of a false lumen. False lumen haemodynamics is a key determinant of aneurysmal growth, rupture, and thrombosis. Quantifying the haemodynamics in the false lumen can provide markers to predict these events and stratify patient risk. While such metrics can be extracted from numerical simulations or imaging modalities such as 4D flow MR, high-resolution experimental data are needed to validate them. The present study provides an in vitro characterization of the flow inside the false lumen of a type B aortic dissection using a patient-specific flexible phantom and Particle Image Velocimetry. A mock circulatory loop imposing patient-specific flow waveforms at the inlet and outlets of the aortic phantom and a refractive index matching blood analog were employed. Time-resolved measurements of the velocity field in four selected planes of the false lumen were acquired. Compared against our previous work on the same dissection assuming rigid walls, the results demonstrate the impact of wall compliance on the flow in the false lumen. They revealed the generation of a jet during the systolic phase that enters the false lumen through the primary tear and impinging on the opposite wall with high velocity, generating a strong rotational flow therein. During the diastolic phase, a reversal of the flow was observed generating multiple vortical structures both inside the true and false lumen. Haemodynamic markers such as false lumen ejection fraction were calculated and compared with clinical measurements. The results provide an insight on AD haemodynamics and highlight the potential of this in vitro method as a validation tool for simulations, as well as to physically test interventions in vitro.

Traumatic brain injury remains a major health concern among civilians and military personnel, with intracranial cavitation hypothesized as a damage mechanism during blunt impacts.

The natural aortic heart valve exhibits an exceptional balance of durability and efficiency, enabling over two billion cycles during a human lifespan. Designing a prosthetic valve that replicates these attributes presents significant challenges. The development of polymeric heart valves offers a promising alternative to existing biologic and mechanical options, aiming to improve durability and hemodynamic performance. This study focuses on the optimized design of the Foldax TRIA polymeric heart valve, leveraging computational modeling to minimize strain energy and enhance structural integrity.

Bicuspid aortic valve (BAV), the most common congenital heart defect, affects ~ 2% of the population but accounts for nearly half of aortic stenosis (AS) cases, typically emerging 10-20 years earlier than in tricuspid aortic valve (TAV) patients. Current tissue-based transcatheter aortic valve replacement (TAVR) devices are designed for TAV anatomy and merely approved for BAV, leading to poor annular fit, paravalvular leak, thrombosis, and limited durability-especially problematic for younger BAV patients. This study presents PolyV-B, a first-of-its-kind polymeric TAVR device tailored to BAV anatomy, aimed at improving durability, hemodynamics, and outcomes through advanced computational modeling.

This study aims to reveal the mechanical microenvironment of trabecular bones during gait by characterizing two key mechanical signals [strain and wall shear stress (WSS)], and to clarify their relationships with the bone volume fraction (BV/TV).

Endothelial-to-mesenchymal transition (EndMT) is the process of endothelial cells undergoing molecular changes that shift their phenotype from that of endothelial cells to that of mesenchymal-like cells. It is a crucial developmental process that has been implicated in various physiological and pathological conditions. EndMT has gained attention as a potential therapeutic target for cardiovascular disease processes, including atherosclerosis, myocardial fibrosis, and vascular calcification. In addition to the assessment of endothelial and mesenchymal markers, the behavioral mechanics of endothelial cells, such as migration and invasion, are often used to identify endothelial cells that have undergone EndMT. However, whether cell chirality may be another mechanobiological marker of EndMT remains unclear. In this study, we aimed to develop an accessible micropatterning platform and created a stereolithography (SLA) 3D printing-based polydimethylsiloxane (PDMS) protein-stamp fabrication platform to create customized patterns of ECM proteins to study endothelial cell chirality during EndMT. Human aortic endothelial cells (HAECs) were treated with the inflammatory cytokine tumor necrosis factor-α (TNF-α), which resulted in the downregulation of the endothelial marker ENOS3 and the upregulation of the mesenchymal markers N-cadherin and transgelin, supporting the induction of EndMT. HAECs were seeded onto fibronectin stripe micropatterns, and cell chirality was measured using custom cell-profiling software. HAECs treated with TNF-α exhibited a shift in cell orientation by approximately 18°, supporting altered cell chirality during TNF-α-induced EndMT. Our work innovates novel methods of studying EndMT by developing a flexible and cost-effective protein-stamp fabrication and image analysis pipeline. This pipeline can be used by researchers to study the endothelial cell chirality in response to EndMT induction.

Accurate identification of gastrointestinal endoscopic anatomical structures is critical for improving diagnostic accuracy and reducing missed detection rates. However, endoscopic image quality may be compromised by various factors including lesion interference and inadequate bowel preparation, while the morphological similarity of certain anatomical structures further complicates recognition in low-quality images. To address these challenges, we propose a Structural Information-Guided Cascaded Feature Fusion Network (SIG-CFFNet). Our approach leverages anatomical prior knowledge to guide the cascaded fusion of CNN and Transformer branch features, while incorporating Depthwise Over-parameterized Convolutional Layer (DO-Conv) to enhance feature representation and computational efficiency during the fusion process. Comprehensive experimental results demonstrate the superior performance of our method across multiple evaluation scenarios: it achieves classification accuracy of 73.47% and 87.05% for normal and pathological endoscopic anatomical structures, respectively; attains 99.61% and 87.83% accuracy on the Kvasir-Capsule and HyperKvasir public datasets; and maintains robust performance with 84.46% and 80.21% accuracy in cross-domain evaluations (COVID19-CT and ISIC2018). Notably, our model demonstrates highly competitive or near state-of-the-art recall rates across multiple test scenarios, confirming its clinical applicability and robustness for real-world implementation.

Developing bone substitute materials that mimic both trabecular and cortical bone remains a major challenge due to the trade-off between bio- and mechano-compatibility, particularly in naturally derived materials. While composites of calcined bone powder and silane-crosslinked alginate exhibit good biocompatibility and mechanical properties resembling those of trabecular bone, their mechanical properties remain insufficient for cortical bone applications.

The study examines the concentration of cryoprotectant (CPA) in an articular cartilage sample during cryopreservation by computing the effective diffusion coefficient using different material models-homogeneous and porous.

Personalized in silico models offer great potential for improving our understanding of atrial arrhythmia mechanisms and for testing therapeutic strategies. However, the level of complexity that can be achieved in these models is often limited by the availability of clinical data. Therefore, evaluating how variations in model detail affect electrophysiological simulation outcomes is essential for evaluating how accurate these models need to be.

Surgical decision-making for Adolescent Idiopathic Scoliosis (AIS) relies on geometrical rather than biomechanical properties, such as the spine's in vivo load characteristics. While both in vivo and in vitro spinal loading experiments can provide valuable insights, a standardized method to compare motion for the Functional Spinal Unit (FSU) is lacking. This work aims to establish a systematic motion parametrization to unambiguously characterize FSU pose changes suitable for a robotic in vivo spinal loading application.

In nasal reconstruction, cartilage struts are used to create the structural foundation of the nose. These struts, carved from autologous or allogeneic grafts of costal cartilage, are often limited by availability, warping, or resorption. This study aimed to investigate the mechanical and biochemical outcomes of applying dynamic bidirectional bending to tissue-engineered constructs, hypothesizing that such stimulation would improve matrix synthesis and mechanical properties.

The neuroprotective effect of hypothermia for mitigation of ischemic and hypoxic damage to the retina is well documented, yet technology to achieve targeted, controlled ocular hypothermia in vivo is lacking. This study evaluated controlled cooling of ocular tissues using a novel scleral contact eye cooler designed to be practical in a clinical setting.

Bicuspid aortic valve (BAV) is a common congenital cardiovascular defect characterized by two, rather than three, cusps. Many BAV patients prematurely develop calcification and aortic stenosis by age 35, which is more severe with fusion of the right and noncoronary (RC/NC) cusps. The mechanisms underlying calcification observed within the coaptation, attachment, and fusion regions of BAV patients are unknown. While abnormal mechanical stimuli induced by the bicuspid anatomy likely plays a role, little is known about regionalized mechanical stimuli in these susceptible cusp regions of young patients prior to calcification. Strongly coupled fluid-structure interaction simulations were conducted using physiologic boundary conditions derived from data of a 23-year-old patient with RC/NC BAV and an age-matched tricuspid aortic valve control. Cusp material properties were implemented for the first time from biaxial testing of non-calcific BAV tissue. Additional simulations elucidated the independent and collective contributions of cusp fusion, material properties and boundary conditions. Results show BAV cusps experience higher time-averaged wall shear stress (TAWSS) over the coaptation region (2.92-fold increase), decreased oscillatory shear index (OSI) within the free edge (1.6-fold) and coaptation (1.4-fold) regions, and an increase in von Mises stress in the coaptation (5.72-fold), belly (6.79-fold), and attachment (5.18-fold) regions of the fused and nonfused cusps. These results reveal putative regions susceptible to calcification in BAV patients experience differences in mechanical stimuli that may contribute to the onset of calcification and provide insight for future in vitro and ex vivo studies focusing on mechanosensitive pathways involved in BAV-induced calcification.

Digital twin (DT) cohorts are collections of models where each member represents an individual real-world asset. DT cohorts can be used for in-silico trials, outlier detection and forecasting, and are used across engineering, industry, and increasingly in personalised medicine. To increase the scalability of DT cohorts, researchers often train emulators to be used as cheap surrogates of computationally expensive mathematical models. Frequently, each cohort member is emulated individually, without reference to other members. We propose that instead, we can treat each DT as a thread in a larger network, and that these threads can be woven together into a digital tapestry using cohort learning methods.

Bone surface registration in current computer-assisted surgical navigation primarily relies on the manual selection of anatomical landmarks and invasive bone surface sampling using mechanical probes. This approach is traumatic, highly dependent on operator expertise, and difficult to automate. Although ultrasound-based registration methods have been explored and demonstrated certain application potential, most existing techniques still rely on manual feature extraction. Moreover, their accuracy is generally limited by the structural and geometric constraints of the probes. To address these limitations, we propose and develop a real-time, noninvasive bone surface depth acquisition system based on A-mode ultrasound, aiming to replace traditional mechanical probing and provide a high-precision, automated, and noninvasive alternative for bone surface registration.