BME frontiers最新文献

Deep learning (DL) models have been widely applied for Alzheimer's disease (AD) stage classification. This scoping review synthesizes recent research to evaluate current performance benchmarks, identify methodological limitations, and highlight translational barriers. DL has potential to augment diagnostic accuracy and accelerate early intervention in AD, but translation requires models that generalize across datasets and integrate into real-world clinical workflows. Following scoping review methodology, 18 peer-reviewed studies published between 2018 and 2024 were analyzed. We extracted dataset sources, preprocessing strategies, model architectures, performance metrics, and translational considerations. Most studies employed convolutional neural networks (CNNs) or transfer learning (TL) backbones with accuracies frequently reported above 90%. Comparative synthesis revealed that TL and custom CNNs achieved similar headline accuracies, with differences of less than one percentage point. Reported performance was highly sensitive to task framing (cross-sectional vs. progression) and dataset provenance, with curated subsets often yielding near-ceiling internal accuracies but limited generalizability. Only one study implemented true external validation, underscoring a critical translational gap. Cost-effectiveness was rarely discussed explicitly; however, several studies indicated that open datasets reduce financial barriers, while adapting pipelines for EMR, or multisite data entails substantial resource demands. DL for AD classification shows consistent high accuracy but limited robustness, with external validation and financial cost-effectiveness remaining underreported. Future progress depends on standardized evaluation protocols, explicit reporting of financial costs, and the development of clinically interpretable, workflow-integrated models.

Objective: This study explores the role of methoxy polyethylene glycol@Elabela-11 (mPEG@ELA-11), a pH-responsive ELA-11 conjugate, in modulating macrophage function and attenuating atherosclerosis, focusing on the protein kinase B (AKT)-mediated endoplasmic reticulum (ER) stress pathway as a molecular target. Impact Statement: We reveal that ELA-11 alleviates atherosclerosis by suppressing macrophage foam cell formation, M1 polarization, and apoptosis via the AKT-ER stress pathway. We also develop mPEG@ELA-11, a novel pH-responsive nanocarrier, to enhance targeted drug delivery and therapeutic efficacy, offering a breakthrough for peptide-based cardiovascular nanomedicine. Introduction: Atherosclerosis, driven by macrophage dysfunction and lipid accumulation, is a major global killer. ELA-11, a fragment of Elabela peptide, shows cardiovascular protective effects, but its role in atherosclerosis and optimal delivery remain unstudied. Methods: Elabela mRNA (APELA) expression was analyzed in human carotid atherosclerotic plaques using real-time quantitative PCR analysis, and serum ELA levels were quantified via enzyme-linked immunosorbent assay in patients with carotid stenosis. In vitro studies on RAW264.7 macrophages evaluated mPEG@ELA-11 effects on oxidized low-density lipoprotein-induced foam cell formation, polarization, and apoptosis. In vivo efficacy was tested in ApoE^-/- mice, comparing mPEG@ELA-11 with free ELA-11, and its pH-responsive release mechanism was characterized. Results: APELA was down-regulated in human atherosclerotic plaques, especially unstable lesions. mPEG@ELA-11 suppressed foam cell formation, M1 polarization, and apoptosis by inhibiting the AKT-ER stress pathway in vitro. In mice, it reduced plaque area more effectively than free ELA-11 attributed to pH-triggered release. Conclusion: The pH-responsive mPEG@ELA-11 alleviates atherosclerosis by modulating macrophages via the AKT-ER stress pathway, with favorable targeting and safety, representing a promising targeted peptide nanomedicine for atherosclerosis.

Objective: Metabolic dysfunction-associated steatotic liver disease (MASLD) is a complex, progressive disorder involving multiple cell types, ranging from simple steatosis to metabolic dysfunction-associated steatohepatitis (MASH), characterized by pro-inflammatory macrophage activation, and can eventually advance to fibrosis, initiated by hepatic stellate cells (HSCs). In vitro multi-cell coculture models are vital tools for elucidating the mechanisms underlying MASLD. Impact Statement: Existing in vitro models for MASLD, including traditional 2-dimensional (2D) cultures and advanced organ-on-a-chip and organoid systems, face challenges in representing multiple cell types and analyzing them individually. Here, utilizing a cell carrier developed in our laboratory, we introduce a series of 3D dynamic coculture models that simulate different stages of MASLD progression and enable individual cell type analysis. Introduction: Currently, no single system provides an optimal balance of control, reproducibility, and analytical convenience. Most in vitro models lack the ability to isolate and analyze individual cell types post-culture, making it difficult to study cell-specific responses in MASLD progression. Methods: The 3D hollow porous sphere cell carrier allows cells to grow on its surface, while the culture device (mini-bioreactor) creates a dynamic environment. The 3 distinct MASLD models were established based on cocultured cell types: steatosis (hepatocytes only), MASH (hepatocytes and macrophages in a 4:1 ratio), and fibrosis (hepatocytes, macrophages, and HSCs in an 8:2:1 ratio). Well-established MASLD mouse models were employed to validate our in vitro 3D dynamic MASLD models, using 7-week-old male C57BL/6J mice fed a high-fat diet. Results: Our models demonstrate a progressive decline in hepatocyte viability and increased lipid accumulation, mirroring in vivo pathology. Additionally, gene expression profiles of our models align with those observed in MASLD-affected mouse livers. Notably, comparative analysis highlights the role of pro-inflammatory macrophages in disrupting hepatocyte lipid metabolism. Conclusion: These models offer a robust platform for investigating MASLD mechanisms and show potential for screening anti-MASLD therapeutics.

Objective: This work aims to construct a functional titanium surface with spontaneous electrical stimulation for immune osteogenesis and antibacteria. Impact Statement: A silver-calcium micro-galvanic cell was engineered on the titanium implant surface to spontaneously generate microcurrents for osteoimmunomodulation and bacteria killing, which provides a promising strategy for the design of a multifunctional electroactive titanium implant. Introduction: Titanium-based implants are usually bioinert, which often leads to inflammation-induced loosening. Electrical stimulation has therapeutic potential; however, its dependence on external devices limits its clinical application. Therefore, designing an electroactive titanium surface with endogenous electrical stimulation capability is a promising strategy to overcome implant failure induced by inflammation. Methods: The silver-calcium micro-galvanic cell was constructed on titanium substrate surfaces by the ion implantation technique. RAW264.7 and MC3T3-E1 were used for cell culture studies with the material to evaluate immunomodulatory and osteogenic abilities of the implant. The expression levels of inflammatory genes and voltage-gated Ca²⁺ channel-related genes were tested for investigating the mechanism of immunoregulation. The antibacterial properties of the modified titanium were assessed. Finally, its immunomodulatory effects in vivo were verified by a mouse subcutaneous inflammation model. Results: The silver-calcium micro-galvanic modified titanium surface generates microcurrents and releases Ca²⁺, which induces macrophage polarization toward the M2 phenotype and promotes osteogenic differentiation via paracrine signaling, exhibiting excellent antibacterial activity. Conclusion: The silver-calcium micro-galvanic cell on titanium could regulate the immune response to promote bone repair and exhibit antibacterial capabilities through noninvasive electrical stimulation, providing a promising strategy for the design of multifunctional electroactive implant surfaces.

Objective: This study characterizes the effects of external conductivity on electroporation to develop methods to overcome potential patient-to-patient variability. Impact Statement: We demonstrate that constant power pulsed electric fields (PEFs) achieve consistent treatment outcomes despite variations in conductivity, thereby improving the predictability and efficacy of electroporation-based therapies. Introduction: Electropermeabilization-based therapies typically deliver static voltages between electrodes to induce cell permeabilization. However, tissue conductivity variations introduce uncertainty in treatment outcomes, as the tissue-specific electric field thresholds that induce electroporation also depend on the extracellular conductivity. Methods: Cell-laden hydrogels were fabricated with varying extracellular conductivities and treated with constant voltage PEFs. The voltages and currents were recorded to calculate the applied powers, and the reversible and irreversible electroporation thresholds were quantified using cell-impermeant and viability assays. Homogeneous and heterogeneous multi-tissue finite element models were employed to simulate the impact of tumor conductivity variability on the outcomes of reversible and irreversible electroporation for constant applied voltage, current, and power PEFs. Additionally, an in vivo murine pancreatic tumor model assessed the correlation between PEF delivery and treatment efficacy. Results: The In vitro experiments revealed that the electric field and current density thresholds were conductivity dependent, whereas the power density thresholds remained stable under variable conductivities. Computational modeling indicated that constant power PEFs best predicted tumor coverage in both homogeneous and heterogeneous multi-tissue models. Similarly, the in vivo tumor responses were also better predicted by applied power rather than voltage or current alone. Conclusions: Applying constant power PEFs enables consistent electroporation outcomes despite variations in conductivity.

Objective: This study engineers leaflet- and 3-dimensional (3D) printing-based implant prototypes for infant mitral valve repair via in vitro cultured mesoangioblasts isolated from the human fetal aorta (AoMAB). Impact Statement: Ultrahigh-molecular-weight polyethylene (UHMWPE) coatings, as well as 3D-printed gelatin methacrylate (GelMA) hydrogels for implants, represent new possibilities for devices used in mitral valve repair. Introduction: Mitral valve prolapse (MVP) repair in pediatric patients is challenging due to somatic growth, patient-prosthesis mismatch, reinterventions, infections, and thromboembolism. Tissue-engineered heart valves (TEHVs) offer potential solutions through conventional and 3D printing biofabrication. Methods: Four materials are evaluated: UHMWPE, UHMWPE coated with polyvinyl alcohol (PVA), UHMWPE coated with PVA and collagen, and 3D-printed GelMA hydrogels. The prototypes are characterized for micro/nanostructural, physicochemical (degradation, contact angle, Fourier transform infrared spectroscopy), and mechanical properties (simple strength tests, dynamic mechanical analysis) and assessed for cytocompatibility using AoMAB cells. A 3D printing mitral valve prototype is analyzed via immunostaining. Results: Results highlight UHMWPE coated with PVA and collagen as the most promising, with degradation (7.30 ± 18.71%), a hydrophilic contact angle (26.13 ± 1.45°), and biocompatibility (177.04 ± 68.92% viability). GelMA prototypes show superior viability (216.77 ± 77.69%) and scalability for 3D printing. Conclusion: UHMWPE coated with PVA and collagen and GelMA demonstrate strong potential for TEHVs, with AoMAB cells facilitating 3D culture and future personalized pediatric applications. Further in vitro validation and thrombogenicity assessments are needed.

Objective: Breast cancer is a common tumor and has a high mortality rate. Gene regulatory networks(GRNs) can genetically facilitate targeted therapies for this disease. Impact Statement: This study proposes a new method to infer GRNs. This new method combining genetic modules and convolutional neural networks is presented to infer GRNs from the RNA sequencing data of breast cancer. Introduction: GRNs play an essential role in many disease treatments. Previous studies showed that GRNs will accelerate tumor therapy. However, most of the existing network inference methods are based on large-scale gene collections, which ignore the characteristics of different tumors. Methods: In this work, weighted gene coexpression network analysis was deployed to screen key genes and gene modules. The gene regulatory associations in gene modules were then transformed into 2-dimensional histogram types. A convolutional neural network was chosen as the main framework to fit the gene regulatory types and infer the GRN. Results: The method integrates genetic data analysis and deep learning perspectives to screen key genes and predict GRNs among key genes. The key genes screened were validated by multiple methods, and the inferred gene regulatory associations were widely validated in real datasets. Conclusion: The method can be used as an auxiliary tool with the potential to predict key genes and the GRNs of key genes. It has the potential to facilitate the therapeutic process and targeted therapy for breast cancer.

Objective: This study aims to explore the therapeutic potential of decellularized adipose matrix (DAM) in rejuvenating photoaged skin by modulating the immune microenvironment. Impact Statement: DAM effectively induces M1 to M2 macrophage polarization and rescues the function of photoaged fibroblasts through paracrine mechanisms, providing a novel strategy for skin antiaging through immune microenvironment remodeling. Introduction: Photoaging, triggered by prolonged ultraviolet exposure, is marked by the depletion of skin structural elements and a persistent inflammatory environment. Current clinical interventions primarily target structural defects, while immune modulation remains underexplored. Therefore, developing biomaterials with both extracellular matrix (ECM) replenishment and immune regulatory functions is crucial for skin regeneration. Methods: A photoaged mouse model was established using ultraviolet B irradiation to validate the inflammatory microenvironment. DAM was prepared via physicochemical decellularization and assessed in vitro for its effects on macrophage polarization and macrophage-fibroblast cross-talk. A DAM-functionalized hyaluronic acid (HA/DAM) hydrogel was developed and evaluated for its effects on skin rejuvenation via subcutaneous injection. Results: In vitro experiments demonstrated that DAM substantially promoted M2 macrophage polarization, and M2-macrophage-conditioned medium further improved fibroblast functions, including oxidative stress resistance, migration, and ECM synthesis. In vivo, HA/DAM hydrogel not only increased dermal thickness and collagen density but also restructured the immune microenvironment through M2 macrophage polarization. Conclusion: DAM offers a novel therapeutic approach for skin rejuvenation by modulating the immune microenvironment, demonstrating notable clinical potential.

Nanotechnology has substantially advanced imaging, therapy, and clinical techniques, playing a crucial role in the development of sustainable functional materials in biomedical engineering. Nanoparticles, used as contrast agents in multimodal imaging, offer notable advantages due to their high surface area-to-volume ratio, enabling functionalization with targeting ligands for improved specificity and sensitivity. They can also carry multiple imaging agents or therapeutic drugs, promoting theranostics, an approach combining diagnosis and treatment. However, the need for high-dose contrast agents raises concerns about nanoparticle toxicity. Green nanotechnology addresses this by developing sustainable nanoparticles through eco-friendly synthesis methods, reducing environmental and health risks. Moreover, by using this method, safer imaging agents that align with current health standards can be generated. In parallel, recent advancements in artificial intelligence (AI) are transforming imaging applications. Beyond simple automation of image interpretation, AI is enhancing image acquisition, management, and interpretation, signaling a future where intelligent systems play a key role in healthcare. This review explores the diverse nanomaterials utilized as contrast agents in multimodal imaging, highlights the importance of green nanotechnology in minimizing toxicity, and emphasizes on the important role of AI in imaging and image-guided therapy. Together, these innovations are advancing precision healthcare, promising a future where diagnostics and treatment are not only more effective but also sustainable.

Objective: This study aims to develop an unsupervised denoising framework for low-dose coronary computed tomography (CT) angiography (LDCTA), which reduces noise while preserving vascular structures. Impact Statement: This work proposes Ves-GAN, a novel denoising framework that meets the challenges of data acquisition and assumptions about noise characteristics. By providing robust noise reduction while maintaining vascular integrity, Ves-GAN facilitates more reliable clinical evaluations and improves the overall quality of cardiovascular diagnosis. Introduction: LDCTA minimizes radiation exposure in cardiovascular imaging but introduces noise and blurring, affecting diagnostic accuracy. Existing denoising methods, such as supervised deep learning models, require paired datasets and rely on noise assumptions. Unsupervised models show promise but often fail to preserve vascular structures, limiting clinical application. Methods: Ves-GAN incorporates a high-frequency-aware data augmentation strategy for robust generalization. The generator employs a high-frequency squeeze-and-excitation module to improve sensitivity to fine vascular features. Additionally, a vessel-consistency loss is introduced to preserve structural integrity during the denoising process. Results: On average, Ves-GAN achieves 7.5% and 10.2% improvements in peak signal-to-noise ratio and structural similarity index metrics compared to existing unsupervised models. Clinical validation involved 50 CT scans reviewed by 3 radiologists, who noted substantial enhancements in vascular clarity and lesion visibility. Conclusion: Ves-GAN outperforms existing unsupervised models in preserving vascular details and noise reduction, significantly enhancing clinical diagnostic reliability.