Bioengineering

24 pages, 55999 KB

Open AccessArticle

A Method for Workout Video Classification via Explainable and Federated Learning

by Ludovica Ciardiello, Patrizia Agnello, Marta Petyx, Fabio Martinelli, Mario Cesarelli, Antonella Santone and Francesco Mercaldo

Bioengineering 2026, 13(6), 603; https://doi.org/10.3390/bioengineering13060603 - 22 May 2026

Abstract

In recent years, the widespread availability of wearable devices and smartphones has enabled the large-scale collection of human activity data, fostering new opportunities for automatic workout recognition and personalized fitness monitoring. However, the centralized storage of video recordings raises critical privacy concerns, particularly [...] Read more.

In recent years, the widespread availability of wearable devices and smartphones has enabled the large-scale collection of human activity data, fostering new opportunities for automatic workout recognition and personalized fitness monitoring. However, the centralized storage of video recordings raises critical privacy concerns, particularly when raw data contain identifiable individuals. Federated Machine Learning provides a paradigm designed with the aim of reducing privacy risks; here, models are collaboratively trained across distributed clients without sharing their sensitive data. In this paper, we propose an approach for workout video classification with Federated Machine Learning, enhanced by explainability through Gradient-weighted Class-Activation Mapping. The proposed method is evaluated on a real-world multi-class exercise video dataset, organized into eight biomechanically coherent macro-classes. In the experimental analysis, we consider several federated configurations in terms of the number of clients, the chosen aggregation strategy, and global communication rounds. The obtained results demonstrate that different aggregation strategies achieve comparable overall accuracy, while explainability effectively highlights the discriminative regions associated with exercise execution, revealing meaningful differences in model behavior between aggregation strategies and uncovering misclassifications driven by contextual biases, demonstrating the trustworthiness of the proposed approach for explainable workout video classification. Full article

(This article belongs to the Special Issue AI and Data Science in Bioengineering: Innovations and Applications)

21 pages, 908 KB

Open AccessArticle

Computer Vision for Movement Observation and Recovery Enhancement (C-MORE): Box and Blocks Test

by Jun Min Kim, Ziqiang (Joe) Zhu, Hari Venugopalan, Vicky Chan, Matthew K. Farrens, Samuel T. King and Andria J. Farrens

Bioengineering 2026, 13(6), 602; https://doi.org/10.3390/bioengineering13060602 - 22 May 2026

Abstract

Stroke is a leading cause of chronic disability, with heterogeneous sensorimotor impairments that are not well captured by standard clinical assessments. While advanced motion capture and robotic systems provide precise measurements, they are not scalable for widespread clinical use. We developed C-MORE (Computer [...] Read more.

Stroke is a leading cause of chronic disability, with heterogeneous sensorimotor impairments that are not well captured by standard clinical assessments. While advanced motion capture and robotic systems provide precise measurements, they are not scalable for widespread clinical use. We developed C-MORE (Computer Vision for Movement Observation and Recovery Enhancement), a smartphone-based framework that uses computer vision and machine learning to automatically score the Box and Blocks Test (BBT) and extract quantitative kinematic metrics. The system combines hand tracking with a custom machine learning (ML) architecture to identify valid block transfers and segment task phases. We evaluated C-MORE in 7 individuals with chronic stroke and a cohort of 10 healthy adults. The system achieved 99.0% agreement with ground-truth scoring, with errors below clinically meaningful thresholds. Kinematic measures derived from the system were sensitive to stroke-related impairments, including reduced movement velocity and increased task duration in affected limbs. Exploratory analyses indicated that grasp-related metrics, particularly the ratio of grasp-to-transfer duration, were correlated with independent measures of proprioception. These findings demonstrate that smartphone-based computer vision can provide accurate, scalable assessment of upper-extremity function. C-MORE offers a practical approach for enhancing clinical evaluation and enabling more precise, individualized rehabilitation strategies. Full article

(This article belongs to the Special Issue Technological Advances in Neurorehabilitation)

29 pages, 1045 KB

Open AccessReview

Experimental Burn Induction in Laboratory Animals: A Scoping Review of Methods, Reproducibility, Operator-Dependent Variability, and Relevance to Soft Tissue Reconstruction and Repair

by Antonios Kyriakopoulos, Michalis Katsimpoulas, Vasilios Kyriakopoulos, Evangelos Felekouras, Stratigoula Sakellariou, Ioannis Kouris and Alexandros Charalabopoulos

Bioengineering 2026, 13(6), 601; https://doi.org/10.3390/bioengineering13060601 - 22 May 2026

Abstract

Background: Experimental animal models remain central to burn research and soft-tissue reconstruction/repair, but method heterogeneity compromises reproducibility, comparability, and translation for depth/area endpoints. Objective: We aimed to map burn-induction methods and examine reproducibility, intentional depth modulation, wound-area stability, validation, and operator-dependent variability. Methods: [...] Read more.

Background: Experimental animal models remain central to burn research and soft-tissue reconstruction/repair, but method heterogeneity compromises reproducibility, comparability, and translation for depth/area endpoints. Objective: We aimed to map burn-induction methods and examine reproducibility, intentional depth modulation, wound-area stability, validation, and operator-dependent variability. Methods: A PRISMA-ScR review, informed by JBI guidance, was conducted without registration but with predefined questions, criteria, and charting domains. PubMed/MEDLINE, Scopus, Web of Science, Embase, and Google Scholar were searched from inception to 30 January 2026. Eligible studies were English peer-reviewed full-text original in vivo animal studies. Two reviewers independently screened records; one charted data, another checked it. Evidence was mapped by modality, exposure-control architecture, validation, and operator-sensitive steps. Results: Studies varied by species, modality, device design, exposure settings, and severity verification. Modalities were contact, scald, steam, and radiant/infrared. Wound area was more reproducible than depth, which depended on temperature, duration, force/pressure, geometry, equilibration, anatomical site, and assessment timing. Histopathology was the main standard, sometimes complemented by morphometry, optical, or perfusion techniques. Operator-sensitive variability involved force, alignment, contact stability, template integrity, exposure geometry, source stability/environmental control. Conclusions: Burn induction is a measurement-system problem; constraining operator-sensitive variables, predefined validation timing, and quantitative variability reporting may improve validity, comparability, and translation. Full article

(This article belongs to the Special Issue Soft Tissue Reconstruction and Repair)

10 pages, 1282 KB

Open AccessArticle

Effects of Flat-Side Design on Torsional and Bending Stress of Nickel–Titanium File by Finite Element Analysis

by Yinjie Yang, Xinfang Cao, Jiwu Zhang, Yuqing Zhou, Songhao Chen and Benxiang Hou

Bioengineering 2026, 13(6), 600; https://doi.org/10.3390/bioengineering13060600 - 22 May 2026

Abstract

Background: This study evaluated the effects of flattening the side of the bending resistance and torsional resistance of nickel–titanium files through finite element analysis of a novel flattened file and a standard nonflattened file. Methods: For torsion analysis, the tip of the file [...] Read more.

Background: This study evaluated the effects of flattening the side of the bending resistance and torsional resistance of nickel–titanium files through finite element analysis of a novel flattened file and a standard nonflattened file. Methods: For torsion analysis, the tip of the file was fixed at 3 mm, generating a torque of 2.5 N·mm at the handle. For bending analysis of curved root canals (45° and 60°), the handle was kept fixed, a force of 1 N was applied at the tip, and the file was fixed at 3 mm. Results: The standard nonflattened file exhibited better torsional resistance. In contrast, the novel flattened file showed improved flexibility under 45° bending. Under this condition, lower maximum von Mises stress was observed in the flattened design compared with the standard file. At 60° bending, stress distribution varied with loading orientation, and higher stress concentrations were observed in the flattened file under specific bending directions, indicating reduced bending resistance under large deformation conditions. Conclusions: Since lateral flattening may reduce the cyclic resistance of files, caution should be exercised in the clinical use of such files. Full article

(This article belongs to the Section Biomedical Engineering and Biomaterials)

► Show Figures

Figure 1

19 pages, 1946 KB

Open AccessArticle

Environment-Driven Synthetic Baseline Analysis and Optimization in Joint Measurement OPM-MEG Arrays

by Wenli Wang, Jianxin Yang, Weinan Xu, Fuzhi Cao, Nan An, Zhenfeng Gao, Min Xiang and Wen Li

Bioengineering 2026, 13(6), 599; https://doi.org/10.3390/bioengineering13060599 - 22 May 2026

Abstract

Optically pumped magnetometer-based magnetoencephalography (OPM-MEG), with its flexible sensor configuration and wide range of application scenarios, has become a powerful complement to conventional superconducting quantum interference device magnetoencephalography (SQUID-MEG). However, this higher flexibility also means that OPM-MEG sensor arrays are more susceptible to [...] Read more.

Optically pumped magnetometer-based magnetoencephalography (OPM-MEG), with its flexible sensor configuration and wide range of application scenarios, has become a powerful complement to conventional superconducting quantum interference device magnetoencephalography (SQUID-MEG). However, this higher flexibility also means that OPM-MEG sensor arrays are more susceptible to interference from complex and variable background magnetic noise. Previous research has shown that deploying reference sensors around the scalp array for noise cancellation is an effective strategy. Nonetheless, the selection of its key parameter, the spatial distance between the reference and scalp sensors, commonly termed the synthetic baseline, predominantly relies on empirical rules and lacks systematic theoretical optimization. To address this issue, this study thoroughly investigates the fundamental impact of the synthetic baseline on the system’s noise suppression performance. Simulation results demonstrate that the optimal baseline length is not a fixed value but varies systematically with environmental noise characteristics and the specific requirements of the source localization task. Building on this analysis, a Baseline Adaptive Reference Optimization (BARO) method is proposed. As an environment-driven strategy, the BARO method automatically determines the optimal baseline configuration by maximizing the output signal-to-noise ratio (SNR). Compared to traditional fixed-baseline configurations, the proposed BARO method significantly enhances the output SNR and effectively reduces the localization error of equivalent current dipoles within the brain across various simulated complex noise scenarios. This work provides a physically interpretable criterion for baseline optimization and offers theoretical support for environment-adaptive configuration of OPM-MEG sensor arrays. Full article

(This article belongs to the Section Biosignal Processing)

► Show Figures

Figure 1

18 pages, 3418 KB

Open AccessArticle

A Brain Connectivity Approach to Detect Diffusion-Weighted Imaging Changes in Post-Traumatic Epilepsy

by Emanuele C. Amato, Claudia Giliberti, Nicola Amoroso, Kseniia Kriukova, Alfonso Monaco, Ester Pantaleo, Tommaso Maggipinto, Loredana Bellantuono, Antonio La Calamita, Roberto Bellotti, Paul M. Vespa, Dominique Duncan and Marianna La Rocca

Bioengineering 2026, 13(6), 598; https://doi.org/10.3390/bioengineering13060598 - 22 May 2026

Abstract

Traumatic brain injury (TBI) is one of the leading causes of acquired epilepsy, with a significant proportion of patients developing post-traumatic epilepsy (PTE) even months or years after the initial injury. The identification of reliable imaging biomarkers able to predict epileptogenesis remains a [...] Read more.

Traumatic brain injury (TBI) is one of the leading causes of acquired epilepsy, with a significant proportion of patients developing post-traumatic epilepsy (PTE) even months or years after the initial injury. The identification of reliable imaging biomarkers able to predict epileptogenesis remains a major clinical challenge. In recent years, diffusion-weighted imaging (DWI) and structural connectome analysis have emerged as promising tools to investigate brain network alterations associated with late seizure development. Machine learning approaches may further support the detection of predictive patterns in complex neuroimaging data. The goal of this study is to perform a binary classification between seizure-free and late seizure-affected patients following TBI, with a specific focus on the identification of the anatomical regions potentially connected with late seizure development. A dataset of 59 diffusion weighted images (DWI) scans from the EpiBioS4Rx project, including 42 seizure-free and 17 late seizure-affected TBI patients, was analyzed. A Random Forest classification algorithm was applied, incorporating network feature importance based on the Gini index to investigate model’s decisions and allow a clinical interpretation. The model reported a 69% ± 0.03 accuracy for discrimination and a 73% AUC ± 0.05. Despite the limited and imbalanced nature of the dataset, and the fact that the performance does not significantly exceed chance once all data-dependent steps are taken into account, our approach allows us to achieve accurate classification results compared to the literature and to identify brain regions potentially associated with epileptogenesis. Full article

(This article belongs to the Special Issue Artificial Intelligence (AI) in Bioengineering: Second Edition)

► Show Figures

Figure 1

17 pages, 1462 KB

Open AccessArticle

Identifying Ovarian Cancer-Associated EV mRNA Expression Profiles Using Unsupervised Machine Learning and Non-Negative Matrix Factorization

by Rama Krishna Thelagathoti, Chao Jiang, Dinesh S. Chandel, Wesley A. Tom, Cleo Sarmiento, Appolinaire Olou, Gary Krzyzanowski and M. Rohan Fernando

Bioengineering 2026, 13(6), 597; https://doi.org/10.3390/bioengineering13060597 - 22 May 2026

Abstract

Extracellular vesicle (EV) transcriptomic data provides a high-dimensional representation of cellular states but remains challenging to interpret due to noise, redundancy, and limited sample sizes. Most existing approaches rely on supervised differential expression analyses, which can be biased and may fail to capture [...] Read more.

Extracellular vesicle (EV) transcriptomic data provides a high-dimensional representation of cellular states but remains challenging to interpret due to noise, redundancy, and limited sample sizes. Most existing approaches rely on supervised differential expression analyses, which can be biased and may fail to capture latent structure in small datasets. In this study, we propose an unsupervised machine learning framework based on non-negative matrix factorization (NMF) to identify latent expression programs from EV mRNA profiles. A structured preprocessing pipeline combining expression filtering, variance selection, ANOVA-based feature selection, and correlation pruning was used to reduce dimensionality and improve signal quality prior to matrix factorization. NMF was applied to decompose the data into interpretable gene modules and sample-specific activation patterns. Model selection was performed using reconstruction error and component stability across multiple initializations. Candidate features were prioritized using a composite ranking score integrating module loadings, group-level expression differences, and model stability. The approach identified a stable low-rank representation capturing dominant patterns in the data and a compact set of informative features. These results demonstrate that unsupervised matrix factorization can effectively extract structured, interpretable signals from small-scale transcriptomic datasets and provide a robust framework for feature prioritization and representation learning in high-dimensional biological data. Full article

(This article belongs to the Special Issue Machine Learning in Precision Oncology: Innovations and Applications)

► Show Figures

Figure 1

14 pages, 3360 KB

Open AccessArticle

First Results Comparing MLC Versus IRIS CyberKnife Collimators in Prostate Stereotactic Body Radiation Therapy in an Italian Oncology Institute

by Gaetano Gagliardo, Marcello Serra, Gianluca Ametrano, Rosario Megna, Valentina d’Alesio, Francesca Buonanno, Cecilia Arrichiello, Rossella Di Franco, Valentina Borzillo, Esmeralda Scipilliti, Rocco Mottareale, Simona Mercogliano, Mariagabriella Pugliese, Maria Quarto, Vincenzo Ravo and Paolo Muto

Bioengineering 2026, 13(6), 596; https://doi.org/10.3390/bioengineering13060596 - 22 May 2026

Abstract

Prostate cancer (PCa) is one of the most common malignancies in men and remains a major cause of cancer-related death worldwide. Radiotherapy is a well-established treatment modality for PCa, offering clinical outcomes comparable to surgical approaches. In recent years, stereotactic body radiotherapy (SBRT), [...] Read more.

Prostate cancer (PCa) is one of the most common malignancies in men and remains a major cause of cancer-related death worldwide. Radiotherapy is a well-established treatment modality for PCa, offering clinical outcomes comparable to surgical approaches. In recent years, stereotactic body radiotherapy (SBRT), characterized by the delivery of high radiation doses in a limited number of fractions, has been increasingly adopted as a standard approach in the treatment of prostate cancer, due to its favorable efficacy and toxicity profile. CyberKnife (CK) is one of the most commonly used hypofractionated radiotherapy techniques. This preliminary study aimed to evaluate and compare the radiation dose delivery and treatment time of CK-based SBRT using two different collimation systems: the multileaf collimator (MLC) and the IRIS variable aperture collimator, a dynamic device that adjusts its opening to simulate different circular field sizes. A total of 19 patients with low-to-intermediate-risk PCa were selected and treated at the Radiation Oncology Department of the National Cancer Institute IRCCS Fondazione G. Pascale in Naples between January 2024 and January 2025. For each patient, two treatment plans were generated—one with the IRIS collimator and one with the MLC. The results demonstrated that the use of the MLC significantly reduced treatment time while maintaining dosimetric quality comparable to IRIS-based plans. These findings support the clinical benefit of MLC implementation in prostate SBRT with the CK system. Full article

(This article belongs to the Special Issue Advanced Systems in Radiotherapy)

► Show Figures

Figure 1

24 pages, 12964 KB

Open AccessArticle

3DAD: Super-Resolution Image Synthesis from Anisotropic CT Images Using a Three-Dimensional Adversarial Diffusion Model

by Jianliang Lu, Ho Ming Cheng, Benjamin Xin Hao Fang, Chun On Anderson Tsang, Sarah Yu, Wai-Kay Seto, Philip Leung Ho Yu and Keith Wan-Hang Chiu

Bioengineering 2026, 13(6), 595; https://doi.org/10.3390/bioengineering13060595 - 22 May 2026

Abstract

High-resolution thin-slice computed tomography (CT) images are often compressed into lower-quality thick-slice images for long-term storage, necessitating synthesis for medical diagnosis. In this paper, we propose a novel 3D adversarial diffusion model (3DAD) for high-fidelity synthesis of thin-slice CT from compressed thick-slice CT. [...] Read more.

High-resolution thin-slice computed tomography (CT) images are often compressed into lower-quality thick-slice images for long-term storage, necessitating synthesis for medical diagnosis. In this paper, we propose a novel 3D adversarial diffusion model (3DAD) for high-fidelity synthesis of thin-slice CT from compressed thick-slice CT. 3DAD is composed of a generator and a discriminator for synthesizing denoised thin-slice images from random noise and source images and distinguishing between noised samples from real and denoised synthetic thin-slice images. Specific models were trained on two-slice to six-slice scenarios for abdominal data, using thick-slice CT compressed from real thin-slice CT as the source. 3DAD was evaluated at the time of HCC diagnosis, at the observation and patient levels, using real thin-slice and synthetic thin-slice CT, with DeLong’s test to compare the similarity of receiver operating characteristic (ROC) curves. We further evaluated 3DAD on real-world data with both thin and thick images, with the synthetic image quality assessed by radiologists and in radiomics feature analysis. Based on the external dataset with 548 samples, the achieved mean squared error (MSE), peak signal-to-noise ratio (PSNR), and structural similarity index measure (SSIM) values were 81.374, 29.478, and 0.916, respectively, for the five-slice scenarios at the portal venous phase. The Areas Under Curves (AUCs) achieved were 0.896 on synthetic thin-slice images compared with 0.889 on real thin-slice images at the observation level (p = 0.028) and 0.854 versus 0.846, correspondingly, at the patient level (p = 0.055). For evaluation on the real-world testing dataset after fine-tuning at the portal venous phase, the MSE, PSNR, and SSIM were 70.435, 30.243, and 0.94, respectively. Radiologist evaluation confirmed the high quality of the synthetic image, with no significant difference in the majority of cases across all five parameters, except for radiologist 2, in realistic and consistent situations, under which at least 41 of 43 synthetic images were assessed as equal to or above grade 3. Our 3DAD enabled the synthesis of thick-slice CT images into high-resolution thin-slice images, facilitating high-fidelity volume image application in HCC diagnosis. Full article

(This article belongs to the Special Issue Artificial Intelligence for Computer-Aided Detection in Biomedical Applications, 2nd Edition)

► Show Figures

Figure 1

19 pages, 5286 KB

Open AccessArticle

The Biomechanical Behavior of Selected Achilles Tendon Revision Constructs: An Exploratory Cadaveric Study

by Horia-Mihnea Fotescu, Dragoș Apostu, Noémi Mosonyi, Daniel Oltean-Dan, Horea Benea, Dan Cosma, Cosmin Cosma and Xavier Martín Oliva

Bioengineering 2026, 13(6), 594; https://doi.org/10.3390/bioengineering13060594 - 22 May 2026

Abstract

Background: Achilles tendon re-rupture following operative repair remains a challenging complication, and biomechanical evidence guiding revision strategies is limited. The mechanical behavior of commonly used revision constructs has not been well characterized. The objective of this exploratory study was to provide a descriptive [...] Read more.

Background: Achilles tendon re-rupture following operative repair remains a challenging complication, and biomechanical evidence guiding revision strategies is limited. The mechanical behavior of commonly used revision constructs has not been well characterized. The objective of this exploratory study was to provide a descriptive biomechanical characterization of commonly used Achilles tendon revision constructs, focusing on viscoelastic behavior, load-to-failure properties, and failure mechanisms under standardized loading conditions. Although limited by the absence of construct replication, this study provides hypothesis-generating biomechanical insight into the failure mechanisms of revision constructs, which may inform future comparative studies and surgical strategy selection. Methods: Four fresh-frozen human cadaveric lower limbs underwent standardized Achilles tendon transection with segmental excision to simulate revision conditions. Five revision techniques were evaluated: tensioned cross-lock Bunnell, Krakow, posterior tibial tendon (PTT) augmentation with Bunnell repair, double Kessler with circumferential running suture, and V–Y advancement combined with three simple sutures and double Kessler. All repairs were performed using No. 2 high-strength suturing by a single surgeon. Constructs underwent stress relaxation testing under a constant 100 N load followed by uniaxial load-to-failure testing. Mechanical parameters and failure modes were recorded. Results: All constructs demonstrated time-dependent stress relaxation. The tensioned cross-lock Bunnell repair retained the highest residual force during sustained loading. The PTT-augmented construct exhibited the highest load to failure among the constructs tested and failed at the tendon substance, whereas non-augmented repairs failed predominantly at the suture–tendon interface. The V–Y advancement construct failed at relatively low applied loads under the applied testing protocol. Conclusions: Achilles tendon revision constructs demonstrate distinct biomechanical behaviors. Augmented constructs exhibited higher resistance to tensile loading in this experimental setting and shifted failure away from the repair site, while non-augmented repairs were limited by suture–tendon interface strength. Given that each construct was tested only once and that one specimen was used sequentially for two repairs, the findings should be interpreted strictly as descriptive and hypothesis-generating, without any basis for comparative or inferential conclusions. Full article

(This article belongs to the Topic Biomedical Engineering, Healthcare and Sustainability, 2nd Edition)

► Show Figures

Figure 1

16 pages, 9283 KB

Open AccessArticle

Wrist-Wearable sEMG Gesture Recognition System Based on ThinNet Lightweight Neural Network

by Zihao Wang, Long Meng, Chen Chen and Hongyu Chen

Bioengineering 2026, 13(6), 593; https://doi.org/10.3390/bioengineering13060593 - 22 May 2026

Abstract

Wearable surface electromyography (sEMG)-based gesture recognition enables intuitive human–machine interaction, but practical deployment is often limited by hardware constraints, model complexity, and inter-subject variability. In this study, we developed a high-performance wrist-worn sEMG acquisition system and a lightweight neural network, ThinNet, to achieve [...] Read more.

Wearable surface electromyography (sEMG)-based gesture recognition enables intuitive human–machine interaction, but practical deployment is often limited by hardware constraints, model complexity, and inter-subject variability. In this study, we developed a high-performance wrist-worn sEMG acquisition system and a lightweight neural network, ThinNet, to achieve efficient and accurate gesture recognition. The wristband features a ring-shaped differential electrode array and embedded filtering modules, achieving a signal-to-noise ratio (SNR) of 66.96 dB, significantly higher than commercial devices. Using data from 100 participants performing six gestures, ThinNet achieved 90.47% inter-subject accuracy, with peak accuracy reaching 96.80% under a three-tier buffered decision strategy. Systematic analysis demonstrated that the model maintains high performance with only 40% fine-tuning data, indicating excellent data efficiency. Importantly, the framework supports scalability across additional users and practical deployment in real-world applications. These results highlight the combined effectiveness of hardware optimization and algorithm design in advancing wearable sEMG-based gesture recognition systems. Full article

(This article belongs to the Special Issue Soft and Flexible Sensors for Biomedical Applications)

► Show Figures

Figure 1

22 pages, 3133 KB

Open AccessArticle

Chitosan-Modified Gold Nanoparticle-Based Electrochemical Immunosensor for C-Reactive Protein Detection

by Bilal Ahmad, Changyun Quan, Xiyue Zhang, Haiyan Xia, Zhenhong Yuan, Chenghua Zhu, Yang Zhang, Haixia Yang, Xueqin Huang, Chunyi Tong, Bin Liu and Binjie Xu

Bioengineering 2026, 13(6), 592; https://doi.org/10.3390/bioengineering13060592 - 22 May 2026

Abstract

C-reactive protein (CRP) is one of the most essential biomarkers for the early detection of inflammation and infection. In this study, we developed a sensitive and selective electrochemical immunosensor for CRP detection, leveraging the unique properties of gold nanoparticles (AuNPs). A nanostructured layer [...] Read more.

C-reactive protein (CRP) is one of the most essential biomarkers for the early detection of inflammation and infection. In this study, we developed a sensitive and selective electrochemical immunosensor for CRP detection, leveraging the unique properties of gold nanoparticles (AuNPs). A nanostructured layer of AuNPs was deposited onto a screen-printed carbon electrode (SPCE), followed by the formation of a self-assembled monolayer (SAM) of L-cysteine and EDC/sulfo-NHS chemistry. The antibody was covalently immobilized onto the modified electrode through optimized dual-crosslinking chemistry. Detection conditions were systematically optimized, with pH 8.0 in Tris buffer providing the best electrochemical response. Electrochemical characterization was performed using cyclic voltammetry (CV), electrochemical impedance spectroscopy (EIS), and differential pulse voltammetry (DPV) in a 5 mM K₃[Fe(CN)₆]/K₄[Fe(CN)₆] redox probe solution containing 0.1 M KCl. CRP detection was achieved by monitoring the increase in charge transfer resistance (Rct) upon specific binding of the target CRP antigen to the immobilized antibody. Spiked recovery experiments showed spiked recovery rates ranging from 98.01% to 107.14%, with a standard deviation below 4%. Regeneration studies demonstrated high efficiency, confirming the suitability of the sensor interface for repeated and reliable measurements. Under optimized conditions, the immunosensor exhibited excellent analytical performance, including a low limit of detection (LOD) of 0.16 µg/mL, a wide linear detection range of 5–100 µg/mL, high selectivity against 13 potential interferents (including inflammatory cytokines), and good reproducibility with a relative standard deviation (RSD) of 3.69%. The sensor also showed strong stability, retaining more than 95% of its signal after 15 days, and high regeneration efficiency of 97% over seven cycles. These results highlight the strong potential of the proposed immunosensor for point-of-care (POC) applications due to its simple fabrication, cost-effectiveness, user accessibility, and robust analytical performance. Full article

(This article belongs to the Special Issue Advancements in Imaging and Sensing of Single Multi-Functional Nanoparticles, Viruses and Organelles)

► Show Figures

Graphical abstract

14 pages, 4296 KB

Open AccessArticle

Artifact-Aware Fungal Detection in Dermatophytosis: A Transformer-Based Approach for KOH Microscopy

by Rana Gursoy, Abdurrahim Yilmaz, Baris Kizilyaprak, Esmahan Caglar, Burak Temelkuran, Huseyin Uvet, Ayse Esra Koku Aksu and Gulsum Gencoglan

Bioengineering 2026, 13(5), 591; https://doi.org/10.3390/bioengineering13050591 - 21 May 2026

Abstract

Dermatophytosis is commonly assessed using potassium hydroxide (KOH) microscopy, yet accurate recognition of fungal hyphae is hindered by preparation-related artifacts, heterogeneous keratin clearance, and notable inter-observer variability. This study presents a transformer-based object detection framework using the RT-DETR architecture for precise, query-driven localisation [...] Read more.

Dermatophytosis is commonly assessed using potassium hydroxide (KOH) microscopy, yet accurate recognition of fungal hyphae is hindered by preparation-related artifacts, heterogeneous keratin clearance, and notable inter-observer variability. This study presents a transformer-based object detection framework using the RT-DETR architecture for precise, query-driven localisation of fungal structures in high-resolution KOH images. A dataset of 2540 routinely acquired microscopy images was manually annotated using a multi-class strategy that explicitly distinguishes fungal elements from confounding artifacts, enabling the model to actively suppress false detections arising from visually similar mimics. To assess architectural trade-offs, RT-DETR was benchmarked against two CNN-based detectors (YOLOv11 and Faster R-CNN) under identical training and inference conditions. Five-fold stratified cross-validation was performed, and each fold-level model was evaluated on the same independent held-out test set (n = 254). Across the five evaluations, RT-DETR achieved a mean AP@0.50 of

89.73 % \pm 1.48 %

, a mean recall of

0.831 \pm 0.011

, and a mean precision of

0.921 \pm 0.014

. At the image level, the model achieved a mean sensitivity of

0.989 \pm 0.022

on the independent test set, with a mean of

0.2 \pm 0.4

missed positive cases across the five evaluations. These results demonstrate the technical feasibility of a transformer-based artificial intelligence (AI) system as a decision-support aid for fungal region detection in KOH microscopy, pending prospective multi-center validation to establish clinical generalisability. Full article

(This article belongs to the Special Issue Biomedical Imaging and Data Analytics for Disease Diagnosis and Treatment, 3rd Edition)

► Show Figures

Figure 1

11 pages, 1254 KB

Open AccessProject Report

GF-Predictability for Dental Implants (GF-PreDImp): A Multidomain Predictive Model for Dental Implant Success—Development, Structure and Clinical Application (Project Report)

by Gustavo Vicentis Oliveira Fernandes, Juliana Campos Hasse Fernandes and Sérgio A. Gehrke

Bioengineering 2026, 13(5), 590; https://doi.org/10.3390/bioengineering13050590 - 21 May 2026

Abstract

Dental implant therapy demonstrates high long-term survival; however, biological, behavioral, and technical complications remain prevalent. The objective of this project report was to introduce GF-Predictability for Dental Implants (GF-PreDImp), a novel, comprehensive pre-surgical multidimensional scoring proposal designed to quantify implant success predictability through [...] Read more.

Dental implant therapy demonstrates high long-term survival; however, biological, behavioral, and technical complications remain prevalent. The objective of this project report was to introduce GF-Predictability for Dental Implants (GF-PreDImp), a novel, comprehensive pre-surgical multidimensional scoring proposal designed to quantify implant success predictability through a structured, evidence-based system. The model integrates six domains, Biological, Behavioral, Hard tissue, Soft tissue, Implant, and Prosthetic, assessing variables into a 100-point composite index. The domains evaluate systemic conditions (20 pts), behavioral habits (20 pts), hard-tissue anatomy (20 pts), soft-tissue characteristics (15 pts), implant parameters (15 pts), and prosthetic/surgical factors (10 pts). The final GF-PreDImp score categorizes predictability into five levels: excellent (≥85), good (70–84), moderate to guarded (55–69), guarded to high risk (40–54), and poor (<40). The tool generates dynamic visual outputs, including radar charts, enabling rapid clinical interpretation. While GF-PreDImp provides a framework for individualized risk stratification, it currently serves as a design proposal. Its implementation can improve clinical decision-making and enhance long-term implant outcomes. Further clinical assessments must be done to confirm the findings in future studies. Full article

(This article belongs to the Special Issue Periodontics and Implant Dentistry—2nd Edition)

► Show Figures

Figure 1

14 pages, 1429 KB

Open AccessArticle

An IoT-Enabled Modular 3D Bioreactor for Vascular Tissue Engineering: Design, Fabrication, and Biological Validation

by Belma Nalbant, Ahmet Ozkurt, Taner Akkan, Tufan Egeli, Thomas Pufe, Zeynep Yuce and Tarkan Unek

Bioengineering 2026, 13(5), 589; https://doi.org/10.3390/bioengineering13050589 - 21 May 2026

Abstract

Three-dimensional (3D) bioreactor systems are essential for vascular tissue engineering as they provide controlled environments that better mimic physiological conditions compared to static culture systems. In this study, an IoT-enabled modular rotating 3D bioreactor platform was designed, fabricated using Fused Deposition Modeling (FDM), [...] Read more.

Three-dimensional (3D) bioreactor systems are essential for vascular tissue engineering as they provide controlled environments that better mimic physiological conditions compared to static culture systems. In this study, an IoT-enabled modular rotating 3D bioreactor platform was designed, fabricated using Fused Deposition Modeling (FDM), and biologically validated. The system integrates a Wi-Fi-supported ESP8266 controller and a touchscreen human–machine interface (HMI), enabling real-time monitoring and remote operation. Agarose-chitosan-based tubular hydrogel constructs were seeded with human aortic smooth muscle cells (HASMCs) and cultured under dynamic conditions for 14 days. Biocompatibility was assessed using a lactate dehydrogenase (LDH) assay, while cellular distribution and mitochondrial activity were evaluated by confocal microscopy using DAPI and MitoTracker staining. Fluorescence intensity was further quantified using ImageJ, and 3D surface plots were generated to visualize spatial signal distribution. The results demonstrated sustained cell viability with decreasing cytotoxicity over time. Confocal analysis confirmed a homogeneous distribution of cells within the hydrogel matrix, and quantitative fluorescence analysis showed significantly higher MitoTracker intensity compared to DAPI, indicating increased metabolic activity under dynamic conditions. These findings suggest that the developed bioreactor provides a stable, controllable, and effective platform for vascular tissue engineering applications. Full article

(This article belongs to the Section Biomedical Engineering and Biomaterials)

► Show Figures

Figure 1

33 pages, 3974 KB

Open AccessArticle

ARTEMIS: An Explainable AI Framework for Multi-Class COVID-19 Diagnosis with a Newly Curated Dataset

by Muhammet Emin Sahin, Hasan Ulutas, Mustafa Fatih Erkoc, Baris Karakaya, Recep Batuhan Günay and Enes Eren Suzgen

Bioengineering 2026, 13(5), 588; https://doi.org/10.3390/bioengineering13050588 - 20 May 2026

Abstract

In this work, we propose ARTEMIS, a novel and highly interpretable deep learning pipeline for the automatic classification of Chest X-ray (CXR) and Computed Tomography (CT) images into different categories related to important clinical outcomes: COVID-19 infection, Community-Acquired Pneumonia (CAP) cases, and Normal [...] Read more.

In this work, we propose ARTEMIS, a novel and highly interpretable deep learning pipeline for the automatic classification of Chest X-ray (CXR) and Computed Tomography (CT) images into different categories related to important clinical outcomes: COVID-19 infection, Community-Acquired Pneumonia (CAP) cases, and Normal cases. Unlike existing models based on the static feature enhancement step, ARTEMIS proposes a learnable preprocessing component that dynamically adapts the image contrast and sharpness in training mode, facilitating adaptive optimization. Our hybrid network combines EfficientNet-B0 backbone with built-in SE attention with the optional lightweight Transformer encoder block to jointly learn local radiological features and global relationships between pixels. Comprehensive experiments have been conducted on five different datasets, which comprise four publicly available ones and one novel CT dataset annotated by radiologists, including X-ray and CT modalities. Experimental results show strong robustness and generalization with macro F1-scores greater than 96% on public datasets and 99.39% accuracy on our new CT dataset. To interpret the decision-making process, Grad-CAM++ is employed to generate class-discriminative saliency maps; the highlighted regions are systematically validated against established radiological criteria by a board-certified radiologist, confirming that model decisions are grounded in clinically meaningful pulmonary findings rather than imaging artifacts. Full article

(This article belongs to the Special Issue Explainable Artificial Intelligence (XAI) in Medical Imaging)

29 pages, 2043 KB

Open AccessReview

Bioengineered Silver Nanoparticles: Next-Generation Biogenic Synthesis Strategies for Precision Biomedical Applications

by Mythileeswari Lakshmikanthan, Sakthivel Muthu and Indra Neel Pulidindi

Bioengineering 2026, 13(5), 587; https://doi.org/10.3390/bioengineering13050587 - 20 May 2026

Abstract

Silver nanoparticles (AgNPs) have attracted substantial scientific interest in biomedical research owing to their unique physicochemical characteristics, broad-spectrum antimicrobial activity, plasmonic properties, and therapeutic versatility. Although conventional physicochemical synthesis methods enable controlled NPs fabrication, their dependence on hazardous reagents, elevated energy input, and [...] Read more.

Silver nanoparticles (AgNPs) have attracted substantial scientific interest in biomedical research owing to their unique physicochemical characteristics, broad-spectrum antimicrobial activity, plasmonic properties, and therapeutic versatility. Although conventional physicochemical synthesis methods enable controlled NPs fabrication, their dependence on hazardous reagents, elevated energy input, and environmentally detrimental processing conditions has stimulated the development of sustainable biogenic alternatives. Biological synthesis utilizing plants, microorganisms, fungi, algae, and purified biomolecules has emerged as an eco-friendly and bio-compatible strategy for AgNP fabrication, enabling simultaneous reduction, stabilization, and intrinsic biofunctionalization of NPs. However, traditional biogenic synthesis remains constrained by limited mechanistic understanding, poor batch reproducibility, inadequate control over physicochemical properties, and challenges in large-scale manufacturing. Recent advances in bioengineering have transformed this field through the integration of metabolic engineering, synthetic biology, microfluidic-assisted synthesis, artificial intelligence-guided process optimization, and continuous-flow biomanufacturing, collectively enabling precision fabrication of biogenic AgNPs with enhanced uniformity, scalability, and functional tunability. Furthermore, strategic surface engineering and functionalization have expanded the applicability of biogenic AgNPs across targeted anticancer therapy, antimicrobial intervention, wound healing, regenerative medicine, drug delivery, and theranostic imaging. Despite these advancements, critical challenges remain regarding nano–bio interactions, toxicological safety, regulatory compliance, and translational scalability. Unlike conventional reviews focused primarily on green synthesis approaches, this review critically highlights emerging bioengineering paradigms that enable programmable, scalable, and precision-controlled biogenic AgNP fabrication. This review comprehensively examines next-generation paradigms and strategies for AgNPs biosynthesis, elucidates the molecular mechanisms governing their formation, highlights emerging functionalization and biomedical application paradigms, and discusses current translational barriers. Forming biogenic composites of AgNPs and heteroatom doped carbon nanodots needs intense research in near future. Full article

(This article belongs to the Section Nanobiotechnology and Biofabrication)

► Show Figures

Figure 1

34 pages, 1526 KB

Open AccessArticle

Robust Multi-Site ADHD Classification via GraphSAGE-Based Functional Connectivity Modeling from rs-fMRI

by Rabab Bousmaha, Khouloud Meribai, Nardjes Bouchemal, Naila Bouchemal and Galina Ivanova

Bioengineering 2026, 13(5), 586; https://doi.org/10.3390/bioengineering13050586 - 20 May 2026

Abstract

Attention Deficit Hyperactivity Disorder (ADHD) is a heterogeneous neurodevelopmental disorder whose diagnosis is mainly based on behavioral assessment and is often delayed due to clinical complexity and limited availability of specialists. Resting-state functional magnetic resonance imaging (rs-fMRI) provides a valuable source of information [...] Read more.

Attention Deficit Hyperactivity Disorder (ADHD) is a heterogeneous neurodevelopmental disorder whose diagnosis is mainly based on behavioral assessment and is often delayed due to clinical complexity and limited availability of specialists. Resting-state functional magnetic resonance imaging (rs-fMRI) provides a valuable source of information for supporting automated and objective diagnosis. However, existing studies often do not fully capture the complex interactions of functional connectivity between different brain regions. To address this limitation, this work proposes a graph-based deep learning framework for ADHD classification from rs-fMRI that combines functional connectivity modeling with graph representation learning. The approach used Phase-Locking Value (PLV)-based connectivity estimation and Graph Sample and Aggregate (GraphSAGE) to jointly capture regional brain activity and inter-regional interactions in a scalable and efficient manner. GraphSAGE improves robustness to noise and inter-subject variability by aggregating information from stable local graph neighborhoods. This integration allows the model to learn discriminative connectivity-aware representations while remaining robust to signal variability and adaptable to multi-site data. The proposed framework was evaluated on the publicly available ADHD-200 dataset across multiple acquisition sites as well as on a combined multi-site dataset. The results indicate consistent performance across individual sites and on the combined dataset. The model achieved an Accuracy of 0.89, an AUC of 0.96, and a Specificity of 0.96 on the combined dataset, outperforming several existing methods in this setting. By integrating PLV-based connectivity with GraphSAGE learning, the approach provides an effective and scalable solution for automated ADHD classification from rs-fMRI data, contributing to data-driven approaches for the analysis of neurodevelopmental disorders. Full article

(This article belongs to the Section Biosignal Processing)

► Show Figures

Graphical abstract

29 pages, 5475 KB

Open AccessArticle

Synergistic Effects of Calcium Phosphate Biomaterials Combined with Honey on Osteochondral Regeneration: A Qualitative Study in an Animal Model

by Pavol Rusnák, Katarína Vdoviaková, Ján Danko, Lenka Krešáková, Filip Humeník, Ľubomír Medvecký, Mária Giretová, Radoslava Štulajterová, Kristína Čurgali, Štefan Tóth, Jozef Bíreš, Filip Korim, Zuzana Čriepoková, Peter Očenáš, Roman Totkovič and Tatiana Špakovská

Bioengineering 2026, 13(5), 585; https://doi.org/10.3390/bioengineering13050585 - 20 May 2026

Abstract

Osteochondral defects of the knee represent a significant clinical challenge due to the limited regenerative capacity of the osteochondral unit. The aim of this study was to evaluate the therapeutic potential of calcium phosphate-based biomaterials combined with honey in a porcine model. Osteochondral [...] Read more.

Osteochondral defects of the knee represent a significant clinical challenge due to the limited regenerative capacity of the osteochondral unit. The aim of this study was to evaluate the therapeutic potential of calcium phosphate-based biomaterials combined with honey in a porcine model. Osteochondral defects were surgically induced and treated with a custom-prepared composite material. Tissue regeneration was assessed using integrated macroscopic and microscopic evaluation, supported by multimodal imaging techniques. The outcomes were compared with both a spontaneous healing group and a control site with native cartilage. The composite biomaterial significantly enhanced osteochondral regeneration, with results comparable to healthy cartilage. Notably, improved structural organization and more advanced healing responses were observed in the treated group compared to spontaneous healing. The beneficial effects are attributed to the anti-inflammatory, antimicrobial, antioxidant, and immunomodulatory properties of honey, which may enhance the regenerative microenvironment and support tissue repair. These findings highlight the potential of calcium phosphate-based biomaterials combined with honey as a promising strategy for osteochondral defect treatment, improving structural, biological, and biomechanical aspects of healing. Full article

(This article belongs to the Special Issue Biomaterials for Cartilage and Bone Tissue Engineering: Third Edition)

► Show Figures

Figure 1

Journal Description

Bioengineering

Latest Articles

Journal Menu

Journal Browser

Highly Accessed Articles

Latest Books

E-Mail Alert

News

Topics

Conferences

Special Issues

Topical Collections

Further Information

Guidelines

MDPI Initiatives

Follow MDPI