Bioengineering最新文献

{"title":"Bridging the Gap Between Experimental and Clinical Applications of Mesenchymal Stromal Cell Therapy in Intervertebral Disc Regeneration.","authors":"Lidija Gradisnik, Borut Prestor, Matic Munda, Tilen Zele, Uros Maver, Nina Kocivnik, Tomaz Velnar","doi":"10.3390/bioengineering13101234","DOIUrl":"10.3390/bioengineering13101234","url":null,"abstract":"<p><p>Intervertebral disc (IVD) degeneration is a major cause of chronic low back pain and disability worldwide. Although current treatment strategies effectively alleviate symptoms, they do not restore disc structure or reverse the underlying biological processes responsible for degeneration. Mesenchymal stromal cell (MSC)-based therapies have emerged as promising regenerative approaches owing to their immunomodulatory, anti-inflammatory, and extracellular matrix-regenerative properties, mediated predominantly through paracrine signalling rather than long-term cellular engraftment. Despite encouraging preclinical results, clinical translation remains limited by poor cell survival within the hostile disc microenvironment, protocol heterogeneity, and insufficient long-term clinical evidence. This review critically examines the current state of MSC-based therapy for IVD degeneration by integrating experimental findings, clinical evidence, and emerging translational strategies. Particular emphasis is placed on the biological mechanisms of MSC-mediated regeneration, comparison of different MSC sources, advances in extracellular vesicle- and exosome-based cell-free therapies, and the growing role of quantitative magnetic resonance imaging, molecular biomarkers, and precision diagnostics in patient selection and treatment monitoring. In addition, the review discusses the major biological, technical, manufacturing, and regulatory challenges that currently limit clinical implementation. Collectively, the evidence indicates that successful translation of MSC-based therapies will require not only optimised cell products but also standardised manufacturing, biomaterial-assisted delivery systems, advanced diagnostic biomarkers, and well-designed multicentre clinical trials. Integrating these advances will be essential for transforming MSC-based therapies into safe, standardised, and disease-modifying treatments for patients with intervertebral disc degeneration.</p>","PeriodicalId":8874,"journal":{"name":"Bioengineering","volume":"13 10","pages":""},"PeriodicalIF":4.4,"publicationDate":"2026-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13701582/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149261584","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Criterion Validity of Wrist-Worn IMU-Derived Smoothness Metrics During Functional Upper-Limb Tasks.","authors":"Serena Cerfoglio, Lucia Donno, Mattia Lovato, Carlalberto Francia, Manuela Galli, Veronica Cimolin","doi":"10.3390/bioengineering13101233","DOIUrl":"10.3390/bioengineering13101233","url":null,"abstract":"<p><p>Wearable inertial measurement units (IMUs) may facilitate quantitative assessment of upper-limb motor function outside laboratory settings; however, the validity of IMU-derived movement smoothness metrics compared with optical motion capture remains unclear. This study aimed to investigate the criterion validity of wrist-worn IMU-derived smoothness metrics during functional upper-limb tasks. Twenty-seven healthy adults (14 females, 13 males) performed two functional tasks, Reaching and Hand-to-mouth, with both upper limbs. Movements were simultaneously recorded using a marker-based motion capture system and two wrist-worn IMUs, both sampling at 100 Hz. Acceleration profiles were obtained from MoCap-derived wrist trajectories and IMU signals, and four smoothness metrics were calculated: Spectral Arc Length (SPARC), normalized root mean square (nRMS), Normalized Jerk (NJ), and Log Dimensionless Jerk (LDLJ). Criterion validity was assessed through Pearson correlation, estimation error (RMSE and MAE), Bland-Altman analysis, and proportional bias assessment. During the Reaching task, IMU-derived metrics showed significant correlations with MoCap values for all indices (r = 0.41-0.74), with NJ showing the strongest association. SPARC showed the smallest mean bias between systems. During the Hand-to-mouth task, significant correlations were observed only for SPARC and NJ (r = 0.39-0.47), with increased discrepancies for nRMS and LDLJ. Wrist-worn IMUs provide comparable smoothness estimates to MoCap during specific upper-limb tasks, but agreement depends on both the selected metric and movement characteristics. These findings highlight the need for task- and metric-specific validation before translating laboratory-based movement quality assessment into wearable settings.</p>","PeriodicalId":8874,"journal":{"name":"Bioengineering","volume":"13 10","pages":""},"PeriodicalIF":4.4,"publicationDate":"2026-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13701579/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149261637","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multispectral PPG Sensing Framework for Noninvasive Glucose Prediction via Multi-Feature Fusion and Micro-Environmental Compensation.","authors":"Lin Huang, Jie Wang, Yuanshuang Fan, Jiansen Lan, Haijun Luo, Rui Wang","doi":"10.3390/bioengineering13101231","DOIUrl":"10.3390/bioengineering13101231","url":null,"abstract":"<p><p>Noninvasive blood glucose prediction using near-infrared spectroscopy remains challenging due to weak glucose-specific absorption, tissue variability, and environmental interference. This study proposes a multispectral photoplethysmography (PPG)-based framework integrating multi-feature fusion, local temperature and humidity compensation, and machine learning regression. The system employs four near-infrared wavelengths (940, 1300, 1550, and 1650 nm) while simultaneously measuring finger contact-site temperature (T) and local relative humidity (H) within the finger clip. A 39-dimensional feature set incorporating optical ratios, pulse-wave morphology, frequency-domain descriptors, participant characteristics, and environmental parameters was constructed. Among six optimized regression models, Extreme Gradient Boosting (XGBoost) achieved the best performance in group-isolated 10-fold cross-validation, with R² = 0.9135, root mean square error (RMSE) = 0.2991, and mean absolute relative difference (MARD) = 2.71%. Clarke error grid analysis (CEGA) showed that 97% of predictions fell within zone A and 100% within zones A and B. Feature ablation demonstrated that local temperature and humidity measurements improved prediction accuracy, with humidity providing a greater improvement than temperature. SHapley Additive exPlanations (SHAP) analysis quantified the contributions of optical, environmental, and participant-level features. A separate session-holdout evaluation yielded R² = 0.7261 and MARD = 6.07%. These findings highlight the potential of the proposed framework for noninvasive glucose prediction. However, the small, predominantly healthy cohort and limited glucose range constrain generalizability, necessitating validation in larger, more diverse populations to establish performance in previously unseen participants and assess clinical applicability.</p>","PeriodicalId":8874,"journal":{"name":"Bioengineering","volume":"13 10","pages":""},"PeriodicalIF":4.4,"publicationDate":"2026-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13701181/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149261687","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Focal Muscle Vibration Combined with Electrical Stimulation for Post-Stroke Upper Limb Spasticity: A Pilot Study.","authors":"Christopher L Pulliam, Terri Z Hisel, Sarah E Grabinski, Victoria C Whitehair, Jayme S Knutson","doi":"10.3390/bioengineering13101228","DOIUrl":"10.3390/bioengineering13101228","url":null,"abstract":"<p><p>Post-stroke upper limb spasticity is a common condition characterized by increased muscle tone which interferes with rehabilitation, limits functional activities, and imposes a significant burden on patients and caregivers. Current treatments, including medications, injections, and physical therapy, often have limited effectiveness or produce side effects. Although some evidence supports focal muscle vibration (FMV) and electrical stimulation individually to reduce spasticity, the benefit of combining these modalities remains unclear. This pilot study explored the immediate effects of combining FMV with electrical stimulation on finger flexor spasticity in stroke survivors, aiming to identify combinations of FMV, neuromuscular electrical stimulation (NMES), and transcutaneous electrical nerve stimulation (TENS) that produce immediate reductions in hypertonia. Five stroke survivors were evaluated in repeated trials (804 trials in total) under 48 different conditions involving FMV (percussive and non-percussive), NMES (35 Hz and 100 Hz), and TENS, applied either alone or in combination to the spastic or antagonist muscles of the forearm. Interventions were applied for 10 s, and the Modified Ashworth Scale (MAS) score was assessed immediately before and after each stimulus. In a multinomial model adjusted for participant, percussive FMV with NMES, particularly at 100 Hz, yielded the greatest immediate reductions in MAS scores, with a higher likelihood of reducing MAS to 0 or 1 than percussive FMV alone; for the 100 Hz combination, this was observed in each of the five participants. In contrast, non-percussive FMV and TENS alone showed limited effectiveness, with a higher probability of no change. These preliminary findings indicate that combining percussive FMV with NMES can produce an immediate reduction in finger flexor hypertonia under the tested conditions. Confirmation in larger, blinded studies with longer stimulation and follow-up periods is needed to establish efficacy.</p>","PeriodicalId":8874,"journal":{"name":"Bioengineering","volume":"13 10","pages":""},"PeriodicalIF":4.4,"publicationDate":"2026-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13701424/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149261629","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"International Regulatory Classification of Minimally Manipulated and Borderline Cell and Tissue Products: Implications for Mesenchymal Stromal Cell-Based Interventions.","authors":"Airat Bilyalov, Nikita Filatov, Natalia Iashina, Andrey Briko, Pavel Chibizov, Irina Nadelyaeva, Alexey Tikhomirov, Anastasia Petrikina, Viatcheslav Vasilyev, Roman Deev, Ilya Eremin, Konstantin Kotenko","doi":"10.3390/bioengineering13101235","DOIUrl":"10.3390/bioengineering13101235","url":null,"abstract":"<p><p>Cell and tissue products used in regenerative medicine are classified differently across jurisdictions according to processing, intended use, and manufacturing setting. This review compares regulatory approaches to products including bone marrow aspirate concentrate, adipose-derived stromal vascular fraction (SVF), microfragmented adipose tissue, nanofat, amniotic products, and cartilage-derived products, with particular attention to their implications for mesenchymal stromal cell (MSC)-based interventions. Legislation, regulatory guidance, enforcement documents, judicial decisions, and scientific publications were examined for the United States, European Union, Russian Federation, Eurasian Economic Union, selected Asian countries, and Arab jurisdictions. In the United States, minimal manipulation and homologous use are central classification criteria. In the European Union, substantial manipulation or use for a different essential function may bring a product within the advanced therapy medicinal product framework. Russian and Eurasian rules require distinction among biomedical cell products, individualized products, medicinal products, transplantation materials, and medical procedures. SVF is a heterogeneous cell population, while microfragmented adipose tissue retains tissue structures. Neither should be equated with an isolated or culture-expanded MSC preparation solely because it contains stromal cells. Regulatory assessment and interpretation of clinical evidence require clear reporting of product composition, processing, claimed mechanism of action, and manufacturing conditions.</p>","PeriodicalId":8874,"journal":{"name":"Bioengineering","volume":"13 10","pages":""},"PeriodicalIF":4.4,"publicationDate":"2026-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13701192/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149261758","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Delay-Programmed Kinetic-Chain Model for Racket Strokes with a Two-Parameter Mechanical Stretch-Shortening-Cycle Extension.","authors":"Sun-Hyun Youn","doi":"10.3390/bioengineering13101232","DOIUrl":"10.3390/bioengineering13101232","url":null,"abstract":"<p><p>Efficient racket acceleration in tennis is commonly attributed to the kinetic chain, but a mechanical model must specify how energy stored at successive joints is released and transferred to the racket. We formulate a minimal planar body-arm-racket model in which each joint stores elastic energy in a latched posture and releases it after an explicit delay. In this framework, optimized delay produces a proximal-to-distal sequence and a superadditive increase in racket-tip speed from 16.6 m s^-1 for one active joint to 30.2 m s^-1 for two joints and 43.1 m s^-1 for three joints, thereby providing a direct mechanical realization of the kinetic chain. The remaining question is how the stretch-shortening cycle (SSC), normally described through complex muscle-tendon physiology, can be incorporated into the same model. We address this by representing SSC action as a joint-level stiffness-modulation block governed by two effective active parameters: a rebound stiffness gain α and a velocity-dependent firing gain β. These parameters are athlete-dependent mechanical descriptors, not universal biological constants. For one illustrative parameter set, SSC engagement raises the peak speed to 46.7 m s^-1 (approximately 8%), but this percentage changes with α and β. Because the size of an individual's active SSC contribution is known to vary substantially from athlete to athlete, this percentage is reported only as an illustration and is not itself the result being claimed. The main contribution is a unified mechanical framework in which time-delayed kinetic-chain release and SSC-like enhancement can be optimized together for athlete-specific posture and stroke design, by adding the two parameters α and β on top of the already delay-optimized passive model without modifying its optimization architecture.</p>","PeriodicalId":8874,"journal":{"name":"Bioengineering","volume":"13 10","pages":""},"PeriodicalIF":4.4,"publicationDate":"2026-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13701319/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149261602","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Hemolysis During Ex Vivo Lung Perfusion Is Associated with Endothelial Injury, Cytokine Release, and Primary Graft Dysfunction.","authors":"Chris H H Chan, Rupa Kanchi, Fernanda C P Mesquita, Patricia Da Costa Lopez, Yaxin Wang, Asma F Farooqui, Wilson Lim, Andres Leon Pena, Emma E Hills, Cristian Coarfa, Abdussalam Elsenousi, Aladdein Mattar, Nirmal S Sharma, Ramiro Fernandez, Camila Hochman-Mendez, Gabriel Loor","doi":"10.3390/bioengineering13101230","DOIUrl":"10.3390/bioengineering13101230","url":null,"abstract":"<p><p>Primary graft dysfunction (PGD) is a major cause of early morbidity and mortality after lung transplantation. During blood-based ex vivo lung perfusion (EVLP), hemolysis releases plasma-free hemoglobin (pfHb) and other erythrocyte-derived products that may promote endothelial injury and inflammation. We investigated whether EVLP-associated hemolysis is associated with cytokine release, endothelial dysfunction, and grade 3 PGD (PGD3). In this prospective, single-center observational study, end-EVLP perfusate from 57 double-lung transplant recipients supported with the Organ Care System Lung was analyzed for pfHb, 26 cytokines, blood counts, and associations with PGD3 within 72 h after transplantation. A mock EVLP blood-loop model was used to evaluate the effects of red blood cell storage duration and circuit run time on hemolysis. Generated hemolytic perfusates were applied to human lung microvascular endothelial cells to assess barrier injury and inflammatory activation. Among the 57 recipients, 32 (56.1%) developed PGD3 at any time point within 72 h after transplantation. Higher end-EVLP pfHb, interleukin-6, and interleukin-8 levels were associated with PGD3, and pfHb correlated positively with interleukin-6 and interleukin-8. Hemolysis increased with longer blood storage and EVLP duration, with storage duration emerging as the dominant contributor in the blood-loop model. Hemolytic perfusate disrupted endothelial junctions, increased permeability, and induced interleukin-6 and interleukin-8 release. EVLP-associated hemolysis is linked to endothelial barrier injury, inflammatory activation, and PGD3 after lung transplantation. Reducing hemolysis through improved blood-product selection and EVLP circuit management may represent a modifiable strategy to improve posttransplant outcomes.</p>","PeriodicalId":8874,"journal":{"name":"Bioengineering","volume":"13 10","pages":""},"PeriodicalIF":4.4,"publicationDate":"2026-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13701639/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149261591","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Nested Evaluation of Sparse Multi-Objective Evolutionary Gene-Panel Discovery for Diabetic Nephropathy.","authors":"Yilei Cong, Tengfei Wu, Xinyu Yang, Yiyun Zhu, Min Tang, Hua Yang, Ping Liu","doi":"10.3390/bioengineering13101229","DOIUrl":"10.3390/bioengineering13101229","url":null,"abstract":"<p><p><b>Background:</b> Diabetic nephropathy gene panels come from small kidney transcriptome cohorts, where screening and adaptive search inflate performance. Evolutionary multi-objective search balances panel size, accuracy, and biological coherence, but dense initialization slows it, and its benefit over ranking or regularized regression is rarely tested under nested evaluation. <b>Methods:</b> Six evolutionary algorithms optimizing leanness, inner balanced accuracy, and Gene Ontology (GO) coherence were evaluated with four repeats of nested five-fold cross-validation on GSE96804. Harmonic initialization was compared with Bernoulli, bounded uniform, and SparseEA initializers; GO was removed or randomized. Comparators were random, top-ranked differentially expressed gene (top-DEG), regularized, and wrapper panels and a permuted-label null. Panels were transferred to GSE30528 and GSE142025. <b>Results:</b> Panels contained 2.10-2.70 genes. Harmonic initialization improved hypervolume over Bernoulli and matched existing sparse initializers. Outer area under the curve (AUC) was 0.973-0.991, comparable to top-DEG and regularized selectors; permuted labels gave 0.510. Real GO raised maximum archive coherence from 0.044 to 0.960, and a GO-aware rule raised representative coherence from 0.098 to 0.620 at equal AUC. External AUC was 0.612-0.726 and 0.849-0.964. <b>Conclusions:</b> Harmonic initialization yields compact Pareto fronts without a size cap, GO guidance supplies coherent panels at equal discrimination, and regularized selection gives the most reproducible panel.</p>","PeriodicalId":8874,"journal":{"name":"Bioengineering","volume":"13 10","pages":""},"PeriodicalIF":4.4,"publicationDate":"2026-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13701386/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149261807","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Use of Bioimpedance Monitoring and Artificial Intelligence in Heart Failure: A Scoping Review.","authors":"Martin Puertas-Albarracín, Luis Giménez-Miranda, Javier Medrano, María Del Carmen Romero-Ternero, Andreea M Oprescu, Alberto Olmo","doi":"10.3390/bioengineering13101227","DOIUrl":"10.3390/bioengineering13101227","url":null,"abstract":"<p><p>Heart failure (HF) is frequently accompanied by congestion and fluid accumulation, which are key drivers of decompensation and hospitalization. Bioimpedance wearable sensors (BI), coupled with artificial intelligence (AI) and clinical decision-support systems (CDSS), offer an emerging opportunity to move HF volume-status assessment beyond conventional clinical monitoring towards earlier risk detection and personalized decision support. This scoping review mapped the available evidence on these integrated approaches. Following PRISMA-ScR guidance, systematic searches were conducted in Scopus, IEEE Xplore, PubMed, and Embase, with the final search performed on August 2, 2026. Ten studies were included, of which six applied AI techniques, two incorporated a CDSS, and two included both AI and CDSS. BI measurements were acquired using heterogeneous approaches, including implanted devices, wearables, and other non-invasive systems, and they encompassed intrathoracic impedance, thoracic BI, or whole-body bioimpedance analysis. AI applications focused on HF decompensation, fluid accumulation, pleural effusion, and the risks of hospitalization or readmission, whereas CDSS primarily provided risk information, alerts, or structured clinical decision support. Overall, the integration of BI sensors with AI or CDSS represents a rapidly developing transformative approach to HF volume-status management, with the potential to enable earlier intervention and more individualized clinical decision-making.</p>","PeriodicalId":8874,"journal":{"name":"Bioengineering","volume":"13 10","pages":""},"PeriodicalIF":4.4,"publicationDate":"2026-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13701678/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149261828","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Generative Artificial Intelligence in Health Research and Care: A Narrative Review of Applications and Implications for Alzheimer's Disease and Related Dementia.","authors":"Sam Omidi, Juanita-Dawne R Bacsu, John Robert Absher, Faezeh Babaieasl, Hadi Kooshiar, S Misagh Imani, Nafiseh Ebrahimi, Matthew Lee Smith, Zahra Rahemi","doi":"10.3390/bioengineering13101225","DOIUrl":"10.3390/bioengineering13101225","url":null,"abstract":"<p><p>The advent of generative artificial intelligence (AI), particularly large language models (LLMs) and deep generative networks, is transforming the landscape of Alzheimer's disease and related dementia (ADRD) research and care. This narrative review aims to synthesize current applications of generative AI in ADRD research and care, highlight ethical and implementation challenges, and propose strategies for responsible integration into clinical practice. We conducted a narrative review of literature, regulatory documents, program reports, and relevant sources published between 2016 and 2026, with a specific focus on cognitive screening, synthetic data generation, digital twin modeling, and caregiver support. Techniques discussed include generative adversarial networks (GANs), variational autoencoders (VAEs), diffusion models, multimodal transformers, and LLM-powered chatbots. Generative AI enables applications such as virtual patient companions, remote cognitive assessments via conversational agents, multimodal image-to-text tools, synthetic data augmentation, disease progression simulation, and improved caregiver communication. Challenges include model explainability, clinical validation, regulatory compliance, privacy, data bias, hallucination risk, patient autonomy, and global disparities. Responsible development requires transparent reporting, stakeholder co-design, rigorous clinical and economic validation, and continuous post-deployment monitoring to support fairness and effectiveness in real-world settings.</p>","PeriodicalId":8874,"journal":{"name":"Bioengineering","volume":"13 10","pages":""},"PeriodicalIF":4.4,"publicationDate":"2026-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13701688/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149261590","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}