Bioengineering最新文献

Artificial intelligence (AI), virtual reality (VR), gamification, and telerehabilitation are increasingly incorporated into neurorehabilitation to deliver adaptive, personalized, and remotely accessible interventions for individuals with stroke and other neurological disorders. These technologies aim to address key limitations in conventional rehabilitation by enhancing training intensity, patient engagement, accessibility, and real-time monitoring. This systematic review synthesizes evidence from clinical and simulation-based studies evaluating AI-assisted systems, non-AI gamified platforms, VR/exergames, telerehabilitation models, and simulation-driven architectures across neurological populations. A comprehensive search of PubMed, Scopus, Embase, CINAHL, and Web of Science (2010-2025) identified randomized controlled trials, pilot and quasi-experimental studies, telerehabilitation systems, VR/exergame interventions, AI-based adaptive tools, and computational or model-driven investigations, guided by a revised PICO framework. Data were extracted using a standardized template, with studies categorized by design, population, technological modality, and outcome domain. Risk of bias was assessed using validated tools, and GRADE was applied to stroke-specific clinical outcomes. Twenty-two studies met the inclusion criteria, encompassing both clinical trials and simulation/modeling research. Clinical studies reported improvements in motor function, balance, gait, swallowing, cognition, and psychosocial well-being, often accompanied by high usability and adherence. AI-enabled systems facilitated adaptive difficulty adjustment, automated feedback, and individualized progression, while non-AI platforms demonstrated strong engagement and meaningful functional gains. Simulation studies provided valuable insights into algorithm behavior, sensor-based modeling, and system optimization. Despite promising multi-domain benefits, methodological heterogeneity, limited long-term follow-up, and inconsistent AI transparency remain key challenges, underscoring the need for standardized outcomes, explainable AI, inclusive design, and robust multicenter trials.

Background: Specific health checkups in Japan aim to prevent and detect non-communicable diseases (NCDs). Lifestyle information and non-invasive measurements obtained during these checkups are valuable for population health monitoring. This study aimed to develop a predictive model for type 2 diabetes mellitus (T2DM) using only non-invasive measurements and to identify key predictors.

Methods: A retrospective observational study was conducted using linked health checkup records and medical claims from a city in Japan. Logistic regression was performed to predict a T2DM diagnosis.

Results: A total of 409 of the 1363 participants were diagnosed with T2DM, including 285 of the 950 participants aged 40-74 years and 124 of the 413 participants aged ≥75 years. The model achieved an area under the receiver operating characteristic curve of 0.680 for those aged 40-74 years and 0.665 for those aged ≥75 years, indicating moderate discrimination. Key predictors included male sex, use of antihypertensive drugs, walking speed, and eating habits within 2 h before bedtime. In particular, male sex, having a slower walking speed, and not eating within 2 h before bedtime were positively associated with T2DM diagnosis. Conversely, the absence of antihypertensive or lipid-lowering medications was negatively associated with T2DM diagnosis.

Conclusion: A model based solely on non-invasive measurements moderately identified individuals at risk for T2DM in this community-based Japanese population. Routinely collected health checkup data may support early identification and targeted preventive strategies.

The integration of 3D printing and artificial intelligence is transforming healthcare management by driving innovations in personalized care, supply chain operations, and clinical workflows. This review offers a comprehensive overview and in-depth analysis of recent (2018-2025) applications where AI technologies enhance 3D printing within healthcare. We explore how AI-powered design and optimization facilitate the creation of patient-specific medical devices, implants, and even bioprinted tissues, while intelligent process control increases both quality and efficiency. Additionally, we examine regulatory and ethical considerations, including the evolution of frameworks for AI-enabled devices, as well as challenges in data governance, validation, and equitable access. The review takes a global perspective, presenting real-world case studies that showcase both successful implementations and ongoing challenges. We also discuss various perspectives and controversies, such as the balance between innovation and safety in autonomous AI design, and highlight areas where further research is needed. In contrast to previous narrative reviews that focus solely on clinical applications or technical aspects, this review uniquely evaluates the combined impact of AI and 3D printing on healthcare management-including cost-effectiveness, governance, decision-making processes, and point-of-care manufacturing. This work is particularly valuable for hospital administrators, clinical operations leaders, health policymakers, and biomedical innovation teams seeking to understand the broader implications of AI-enhanced 3D printing in healthcare management. Nevertheless, despite promising advancements, the field is constrained by heterogeneous evidence, a lack of standardized evaluation metrics, and insufficient long-term outcome data, which together limit the ability to fully assess the sustained impact of AI-integrated 3D printing in healthcare environments.

Prosthetic hands have seen significant improvements in recent years, enabling increasingly more natural interactions between patients with upper limb loss and their environment. Nonetheless, progress is continuously being made to enhance user acceptance, which remains a major drawback in such systems. The efficiency of the actuation mechanism is a critical parameter when designing these devices. Maximising actuation approach efficiency enables the use of smaller and lighter motors, thus decreasing the overall weight of the solution. Simultaneously, increased efficiency contributes to more precise motor control. Within this context, the present work introduces a novel actuation concept. Conventional tendon-pulley mechanisms are often susceptible to tendon slippage; therefore, a hobbed tendon-pulley approach was investigated to maintain cable tension more consistently and efficiently. This approach aims to provide smoother operation, improved reliability, and a reduced risk of mechanical failure due to tendon slippage. Simultaneously, the capability of holding and maintaining a set force is of utmost importance in these systems, and the force-feedback system is usually a major concern. The present work also focuses on comparing current and pressure-based control methodologies for the developed prosthesis. The current-based approach had the significant advantage of not requiring external sensors to be assembled on the prosthesis and not relying on the point of application of force being inside the sensor's active area. During these tests, the prosthesis successfully grasped various objects of different sizes, shapes, stiffnesses, and weights using a current-based approach, without observable tendon slippage.

Three-dimensional (3D) anatomic modeling derived from high-resolution medical imaging, such as computed tomography (CT) and magnetic resonance imaging (MRI), has been increasingly adopted in preclinical testing and device development. This white paper describes a cardiac-specific workflow that integrates 3D printing and silicone molding for support device development and procedural simulation. Patient-derived computed tomography angiography data were segmented using FDA-cleared medical modeling software to isolate the left ventricular anatomy and were further processed in computer-aided design (CAD) to ensure accurate physiological wall thickness and structural fidelity. Material jetting 3D printing was performed on a Stratasys J750 using material distributions designed to mimic the mechanical properties of myocardium, thereby approximating myocardial compliance. In parallel, stereolithography apparatus molds were designed from the left ventricle CAD model to cast transparent, pliable left ventricular models in Sorta-Clear™ 18 silicone. The 3D-printed models preserved intricate morphological detail and were suitable for mechanical manipulation and device deployment studies, whereas silicone models offered tunable mechanical properties, transparency for visualization, and durability for repeated use. Together, these complementary modalities provided rapid manufacturing capability and application-relevant physical representation. Case-specific parameters, strengths, and limitations of both models in enhancing patient care and device testing are highlighted, with relevance to heart failure applications. Current knowledge gaps, workflow and integration challenges, and future opportunities are identified, positioning this work as a reference framework for continued innovation in anatomic modeling. Within the collaborative framework of Mayo Clinic's Anatomic Modeling Unit and Simulation Center, this integrated modeling workflow demonstrates the value of multidisciplinary collaboration between engineers and clinicians. Clinically, these patient-specific left ventricular models may enable pre-procedural device sizing and positioning and may support simulation of mechanical circulatory support (MCS) deployment while identifying possible anatomic constraints prior to intervention. This workflow has direct applicability in advanced heart failure patients undergoing MCS support, such as the Impella axillary MCS device or the durable LVAD, with potential to reduce procedural uncertainty while reducing complications and improving peri-procedural outcomes. Additionally, these models also serve as high-accuracy educational tools, enabling trainees and multidisciplinary care teams to visualize and possibly rehearse procedural steps while gaining hands-on experience in a risk-free environment.

Abdominal Aortic Aneurysm (AAA) remains a significant public health challenge, with an 82.1% increase in related fatalities from 1990 to 2019. In the United States alone, AAA complications resulted in an estimated 13,640 deaths between 2018 and 2021. In clinical practice, computed tomography angiography (CTA) is the primary imaging modality for monitoring and pre-surgical planning of AAA patients. CTA provides high-resolution vascular imaging, enabling detailed assessments of aneurysm morphology and informing critical clinical decisions. However, manual segmentation of CTA images is labor-intensive and time consuming, underscoring the need for automated segmentation algorithms, particularly when feature extraction from clinical images can inform treatment decisions. We propose a framework to automatically segment the outer wall of the abdominal aorta from CTA images and estimate AAA wall stress. Our approach employs a patch-based dilated modified U-Net model to accurately delineate the outer wall boundary of AAAs and Nonlinear Elastic Membrane Analysis (NEMA) to estimate their wall stress. We further integrate Non-Uniform Rational B-Splines (NURBS) to refine the segmentation. During prediction, our deep learning architecture requires 17±0.02 milliseconds per frame to generate the final segmented output. The latter is used to provide critical insight into the biomechanical state of stress of an AAA. This modeling strategy merges advanced deep learning architecture, the precision of NURBS, and the advantages of NEMA to deliver a robust and efficient method for computational analysis of AAAs.

The anatomy and mechanical strength of aortic valve leaflets are critical determinants of their biomechanical behavior and long-term structural integrity. The embryonic developmental period, when valves are forming, is critical to establish baseline leaflet properties. However, fetal stages of valve development, when valve leaflets are still forming and remodeling, are not well understood. The goal of this study is to investigate the biomechanical stress and deformation modes of developing valve leaflets during systole, and how leaflet biomechanics are affected by anatomy and material properties. To this end, the study employs a parametric approach to model the leaflet anatomy of an HH40 chick embryo, used here as a model of fetal cardiac development. To perform biomechanical analysis, a pressure profile derived from in ovo Doppler ultrasound measurements was applied, and an Ogden hyperelastic material model was employed following a sensitivity analysis. To determine the effect of valve anatomy on leaflet tissue deformation and stresses, we changed the leaflet midline curve (belly curve) from its native curvature to a linear profile and quantified biomechanical responses. Our analysis revealed a strong decrease in average leaflet effective stress as the belly curvature was shifted towards a linear profile. However, this reduction in average stress was at the expense of a biomechanical trade-off. The shift induced a progressive localization of stress concentration at the leaflet tips and commissures, and a distinct bending deformation mode at the tip under peak load. Our findings demonstrate that while the belly curve of the leaflet modulates tissue stress during valve opening, a low-stress anatomy does not align with hemodynamic performance. This work characterizes competing leaflet biomechanical responses (stress reduction versus failure modes) that shape valve leaflet formation, providing fundamental insights into developmental valve biomechanics.

Immune checkpoint inhibitors (ICIs) targeting PD-1/PD-L1 and CTLA-4 are widely used in the treatment of several cancers and have significantly improved survival outcomes in responsive patients. However, a substantial proportion of patients fail to benefit from these therapies, underscoring the urgent need for accurate prediction of ICI response. We propose a deep learning framework, ICIsc, to accurately predict ICI response by integrating single-cell RNA sequencing (scRNA-seq) data with protein large language models. Specifically, patient representations are constructed using transcriptomic profiles and immune-related gene set scores as latent embedding features, while drug representations are derived from amino acid sequences of ICI encoded by the Evolutionary Scale Modeling 2 (ESM2). For bulk data, ICIsc employs a bilinear attention module to fuse patient and drug embeddings for response prediction. For scRNA-seq data, ICIsc infers cell-cell interactions using a single-sample network (SSN) approach and applies GATv2 to model immune microenvironment heterogeneity at the single-cell level. Benchmark evaluations and independent validation demonstrate that ICIsc consistently outperforms baseline models and exhibits robust generalization performance. SHAP-based interpretability analysis further identifies key genes (e.g., GAPDH) associated with immunotherapy response and patient prognosis. Overall, ICIsc provides an accurate and interpretable framework for predicting immunotherapy outcomes and elucidating underlying mechanisms.

Carotid plaque classification based on ultrasound echogenicity and quantification of plaque burden are crucial in stroke risk assessment. In this work, we propose a framework that leverages the synergy between classification and segmentation by sharing plaque location information to enhance the performance of both tasks. Our cascaded framework integrates a ResNet-based classifier (Masked-ResNet-DS) with MedSAM, a medically adapted version of the Segment Anything Model for joint classification and segmentation of carotid plaques from 2D ultrasound images. Ground truth boundaries are used to guide region-specific feature pooling in the classifier, helping it focus on plaques during training. Since ground truth boundaries are unavailable at inference, we introduce a two-iteration strategy: the first generates a class activation map (CAM), which is then used for focused pooling in the second iteration to predict plaque type. The CAM is also used as a prompt to guide MedSAM for segmentation. To ensure accurate localization, the CAM is supervised during training using a Dice loss against the segmentation ground truth. Masked-ResNet-DS achieves a mean F1-score of 96.7% in plaque classification, at least 3.2% higher than competing methods. Ablation studies confirm that ground truth-based pooling and CAM supervision both improve classification. CAM-guided MedSAM achieves a Dice similarity coefficient (DSC) of 86.6%, outperforming U-Net and nnU-Net by 5.9% and 3.6%, respectively. In addition, CAM prompts improve MedSAM's DSC by 2.2%. By sharing plaque location between classification and segmentation, the proposed method improves both tasks and provides a more accurate tool for stroke risk stratification.

Retinal transplantation offers promise for restoring vision in advanced retinal degeneration. However, manual delivery of retinal sheets is often hindered by imprecise placement and collateral tissue damage resulting from instrument instability. We introduce a novel, 3D-printed, motorized retinal sheet injector designed to enhance placement accuracy and minimize tissue injury. The motorized injector features an Arduino-controlled foot pedal with three discrete actuator positions ("Min", "Mid", "Max"). When compared via frame-by-frame motion analysis, the motorized system reduced tip variance by approximately threefold over manual methods. In addition, in in vitro gelatin trials, the motorized injector achieved significantly higher placement accuracy versus the manual injector, which suffered from occasional complete misplacements. The novel motorized retinal sheet injector markedly improves stability and placement accuracy relative to manual methods, potentially reducing complications associated with subretinal delivery. Safer subretinal delivery can pave the way for innovative research and advanced treatment for retinal disease.