Bioengineering

39 pages, 30587 KiB

Open AccessArticle

Hierarchical Swin Transformer Ensemble with Explainable AI for Robust and Decentralized Breast Cancer Diagnosis

by Md. Redwan Ahmed, Hamdadur Rahman, Zishad Hossain Limon, Md Ismail Hossain Siddiqui, Mahbub Alam Khan, Al Shahriar Uddin Khondakar Pranta, Rezaul Haque, S M Masfequier Rahman Swapno, Young-Im Cho and Mohamed S. Abdallah

Bioengineering 2025, 12(6), 651; https://doi.org/10.3390/bioengineering12060651 - 13 Jun 2025

Abstract

Early and accurate detection of breast cancer is essential for reducing mortality rates and improving clinical outcomes. However, deep learning (DL) models used in healthcare face significant challenges, including concerns about data privacy, domain-specific overfitting, and limited interpretability. To address these issues, we [...] Read more.

Early and accurate detection of breast cancer is essential for reducing mortality rates and improving clinical outcomes. However, deep learning (DL) models used in healthcare face significant challenges, including concerns about data privacy, domain-specific overfitting, and limited interpretability. To address these issues, we propose BreastSwinFedNetX, a federated learning (FL)-enabled ensemble system that combines four hierarchical variants of the Swin Transformer (Tiny, Small, Base, and Large) with a Random Forest (RF) meta-learner. By utilizing FL, our approach ensures collaborative model training across decentralized and institution-specific datasets while preserving data locality and preventing raw patient data exposure. The model exhibits strong generalization and performs exceptionally well across five benchmark datasets—BreakHis, BUSI, INbreast, CBIS-DDSM, and a Combined dataset—achieving an F1 score of 99.34% on BreakHis, a PR AUC of 98.89% on INbreast, and a Matthews Correlation Coefficient (MCC) of 99.61% on the Combined dataset. To enhance transparency and clinical adoption, we incorporate explainable AI (XAI) through Grad-CAM, which highlights class-discriminative features. Additionally, we deploy the model in a real-time web application that supports uncertainty-aware predictions and clinician interaction and ensures compliance with GDPR and HIPAA through secure federated deployment. Extensive ablation studies and paired statistical analyses further confirm the significance and robustness of each architectural component. By integrating transformer-based architectures, secure collaborative training, and explainable outputs, BreastSwinFedNetX provides a scalable and trustworthy AI solution for real-world breast cancer diagnostics. Full article

(This article belongs to the Special Issue Breast Cancer: From Precision Medicine to Diagnostics)

► Show Figures

Figure 1

23 pages, 903 KiB

Open AccessReview

OCT in Oncology and Precision Medicine: From Nanoparticles to Advanced Technologies and AI

by Sanam Daneshpour Moghadam, Bogdan Maris, Ali Mokhtari, Claudia Daffara and Paolo Fiorini

Bioengineering 2025, 12(6), 650; https://doi.org/10.3390/bioengineering12060650 - 13 Jun 2025

Abstract

Optical Coherence Tomography (OCT) is a relatively new medical imaging device that provides high-resolution and real-time visualization of biological tissues. Initially designed for ophthalmology, OCT is now being applied in other types of pathologies, like cancer diagnosis. This review highlights its impact on [...] Read more.

Optical Coherence Tomography (OCT) is a relatively new medical imaging device that provides high-resolution and real-time visualization of biological tissues. Initially designed for ophthalmology, OCT is now being applied in other types of pathologies, like cancer diagnosis. This review highlights its impact on disease diagnosis, biopsy guidance, and treatment monitoring. Despite its advantages, OCT has limitations, particularly in tissue penetration and differentiating between malignant and benign lesions. To overcome these challenges, the integration of nanoparticles has emerged as a transformative approach, which significantly enhances contrast and tumor vascularization at the molecular level. Gold and superparamagnetic iron oxide nanoparticles, for instance, have demonstrated great potential in increasing OCT’s diagnostic accuracy through enhanced optical scattering and targeted biomarker detection. Beyond these innovations, integrating OCT with multimodal imaging methods, including magnetic resonance imaging (MRI), positron emission tomography (PET), and ultrasound, offers a more comprehensive approach to disease assessment, particularly in oncology. Additionally, advances in artificial intelligence (AI) and biosensors have further expanded OCT’s capabilities, enabling real-time tumor characterization and optimizing surgical precision. However, despite these advancements, clinical adoption still faces several hurdles. Issues related to nanoparticle biocompatibility, regulatory approvals, and standardization need to be addressed. Moving forward, research should focus on refining nanoparticle technology, improving AI-driven image analysis, and ensuring broader accessibility to OCT-guided diagnostics. By tackling these challenges, OCT could become an essential tool in precision medicine, facilitating early disease detection, real-time monitoring, and personalized treatment for improved patient outcomes. Full article

(This article belongs to the Special Issue Biomedical Applications of Optical Coherence Tomography, Second Edition)

► Show Figures

Figure 1

26 pages, 1080 KiB

Open AccessReview

Toward Integrative Biomechanical Models of Osteochondral Tissues: A Multilayered Perspective

by Bruna Silva, Marco Domingos, Sandra Amado, Juliana R. Dias, Paula Pascoal-Faria, Ana C. Maurício and Nuno Alves

Bioengineering 2025, 12(6), 649; https://doi.org/10.3390/bioengineering12060649 - 13 Jun 2025

Abstract

Understanding the complex mechanical behavior of osteochondral tissues in silico is essential for improving experimental models and advancing research in joint health and degeneration. This review provides a comprehensive analysis of the constitutive models currently used to represent the different layers of the [...] Read more.

Understanding the complex mechanical behavior of osteochondral tissues in silico is essential for improving experimental models and advancing research in joint health and degeneration. This review provides a comprehensive analysis of the constitutive models currently used to represent the different layers of the osteochondral region, from articular cartilage to subchondral bone, including intermediate regions such as the tidemark and the calcified cartilage layer. Each layer exhibits unique structural and mechanical properties, necessitating a layer-specific modeling approach. Through critical comparison of existing mathematical models, the viscoelastic model is suggested as a pragmatic starting point for modeling articular cartilage zones, the tidemark, and the calcified cartilage layer, as it captures essential time-dependent behaviors such as creep and stress relaxation while ensuring computational efficiency for initial coupling studies. On the other hand, a linear elastic model was identified as an optimal starting point for both the subchondral bone plate and the subchondral trabecular bone, reflecting their dense and stiff nature, and providing a coherent framework for early-stage multilayer integration. This layered modeling approach enables the development of physiologically coherent and computationally efficient representations of osteochondral region modeling. Furthermore, by establishing a layer-specific modeling approach, this review paves the way for modular in silico simulations through the coupling of computational models. Such an integrative framework supports scaffold design, in vitro experimentation, preclinical validation, and the mechanobiological exploration of osteochondral degeneration and repair. These efforts are essential for deepening our understanding of tissue responses under both physiological and pathological conditions. Ultimately, this work provides a robust theoretical foundation for future in silico and in vitro studies aimed at advancing osteochondral tissue regeneration strategies. Full article

(This article belongs to the Section Biomechanics and Sports Medicine)

► Show Figures

Figure 1

12 pages, 907 KiB

Open AccessArticle

Development and Evaluation of a 3D Motion Capture Model for Upper Extremity Kinematics During Wheelchair Maneuvering in Individuals with Spinal Cord Injuries: A Pilot Study

by Lina Bunketorp Käll, Gudni Rafn Harðarson, Erik Tullin, Ann-Sofi Lamberg, Roy Tranberg and Johanna Wangdell

Bioengineering 2025, 12(6), 648; https://doi.org/10.3390/bioengineering12060648 - 12 Jun 2025

Abstract

Spinal cord injury (SCI) often necessitates the use of a manual wheelchair, which can overload the shoulders and contribute to upper extremity (UE) pain. Currently, no standardized methods exist to assess UE kinematics during wheelchair propulsion. This study aimed to develop and evaluate [...] Read more.

Spinal cord injury (SCI) often necessitates the use of a manual wheelchair, which can overload the shoulders and contribute to upper extremity (UE) pain. Currently, no standardized methods exist to assess UE kinematics during wheelchair propulsion. This study aimed to develop and evaluate a marker-based motion capture model for analyzing UE movement during wheelchair use, with a secondary goal of assessing test–retest reliability. The study was conducted in two phases: (1) development of the motion analysis model and (2) reliability testing. Eleven participants with SCI were included. Reliability was assessed using intraclass correlation coefficients (ICCs) across 15 movement parameters, including total range of motion and minimum and maximum movement values. The model demonstrated good test–retest reliability. For minimum movement, 12 of 15 parameters were significant (ICC = 0.681–0.965). For maximum movement, 13 of 15 were significant (ICC = 0.726–0.981). For total range of motion, 12 of 15 showed significant reliability (ICC = 0.596–0.952). In conclusion, the motion capture model showed promising reliability for assessing UE kinematics during wheelchair maneuvering in individuals with SCI. However, due to the small sample size, further research is needed to validate and refine the model. Full article

(This article belongs to the Special Issue Advances in Physical Therapy and Rehabilitation)

► Show Figures

Figure 1

13 pages, 410 KiB

Open AccessReview

Steal Syndrome in Free Flap Microvascular Reconstruction of the Lower Extremity: Systematic Review of Incidence, Risk Factors, and Surgical Management

by Georgios Karamitros, Ilias Iliadis, Raymond A. Pensy and Gregory A. Lamaris

Bioengineering 2025, 12(6), 647; https://doi.org/10.3390/bioengineering12060647 - 12 Jun 2025

Abstract

Background: Steal syndrome in the setting of microvascular reconstruction refers to a phenomenon whereby blood flow is diverted from the native tissue to the free flap, leading to ischemia and potential limb loss. In the present study, we aim to comprehensively evaluate [...] Read more.

Background: Steal syndrome in the setting of microvascular reconstruction refers to a phenomenon whereby blood flow is diverted from the native tissue to the free flap, leading to ischemia and potential limb loss. In the present study, we aim to comprehensively evaluate the occurrence and management of steal syndrome in free flap reconstruction of the lower extremities. Methods: A thorough literature search was conducted across the MEDLINE, Embase, Cochrane Library, and Scopus databases up to 29 January 2025. Studies were selected based on predefined inclusion criteria focusing on free flap microvascular reconstruction in the lower extremities with a focus on steal syndrome. Two independent reviewers assessed and extracted data. Results: Three studies were included, involving seven patients, with a mean age of 65.66 ± 5.89 years, who developed steal syndrome following free flap microvascular reconstruction. The most common revision involved below-the-knee amputation (BKA) due to ischemic complications. Comorbidities such as peripheral vascular disease (PVD), diabetes, and hypertension were present in all cases. The majority of anastomoses (85.7%) were end-to-side (ETS), with only one case utilizing a flow-through configuration. The majority of cases (n = 5, 71.4%) were reconstructed using latissimus dorsi (LD) flaps, with the remaining two cases using rectus abdominis (n = 1) and gracilis (n = 1) flaps. The recipient vessel was the anterior tibial artery in two patients (28.6%), the dorsalis pedis artery in two patients (28.6%), and the popliteal artery in three patients (42.9%). The most common salvage procedure was below-the-knee amputation (BKA), performed in four patients (57.1%). One patient required revision of the venous anastomosis and flap debridement, followed by a Chopart amputation (n = 1, 14.3%). Conclusions: The occurrence of steal syndrome in free flap microvascular reconstruction of the lower extremities is rare but can lead to significant complications, including amputation. The findings indicate that steal syndrome is more likely in patients with pre-existing vascular conditions such as PVD and diabetes. While surgical technique and flap type may influence its development, further studies are needed to identify specific anatomical and clinical predictors. The absence of a unified treatment guideline underscores the need for further investigation into effective management strategies to prevent amputation and optimize patient outcomes. Full article

(This article belongs to the Special Issue Surgical Wound Infections and Management)

► Show Figures

Figure 1

14 pages, 1327 KiB

Open AccessArticle

Exploration of Cytokines That Impact the Therapeutic Efficacy of Mesenchymal Stem Cells in Alzheimer’s Disease

by Herui Wang, Chonglin Zhong, Yi Mi, Guo Li, Chenliang Zhang, Yaoyao Chen, Xin Li, Yongjun Liu and Guangyang Liu

Bioengineering 2025, 12(6), 646; https://doi.org/10.3390/bioengineering12060646 - 12 Jun 2025

Abstract

Current therapies for Alzheimer’s disease (AD) includes acetylcholinesterase inhibitors, NMDA receptor antagonists, and amyloid beta (Aβ)/Tau-targeting drugs. While these drugs improve cognitive decline and target the pathological mechanisms, their outcomes still are still in debate. Mesenchymal stem cells (MSCs) offer a regenerative approach [...] Read more.

Current therapies for Alzheimer’s disease (AD) includes acetylcholinesterase inhibitors, NMDA receptor antagonists, and amyloid beta (Aβ)/Tau-targeting drugs. While these drugs improve cognitive decline and target the pathological mechanisms, their outcomes still are still in debate. Mesenchymal stem cells (MSCs) offer a regenerative approach by modulating neuroinflammation and promoting neuroprotection. Although the paracrine of MSCs is efficient in various AD preclinical studies and the exosomes of MSCs have entered clinical trials, the key cytokines driving the efficacy remain unclear. Here, we evaluated human umbilical cord-derived MSCs (hUC-MSCs) and employed gene-silenced MSCs (siHGF-MSCs, siTNFR1-MSCs, siBDNF-MSCs) in APP/PS1 AD mice to investigate specific mechanisms. hUC-MSCs significantly reduced Aβ/Tau pathology and neuroinflammation, with cytokine-specific contributions: silencing HGF predominantly reduced Aβ/Tau clearance, although silencing TNFR1 or BDNF showed modest effects; silencing TNFR1 or BDNF more prominently weakened anti-neuroinflammation, while silencing HGF exerted a weaker influence. All three cytokines partially contributed to oxidative stress reduction and cognitive improvements. Our study highlights MSC-driven AD alleviation as a multifactorial strategy and reveals specific cytokines alleviating different aspects of AD pathology. Full article

(This article belongs to the Special Issue Nerve Regeneration)

► Show Figures

Graphical abstract

15 pages, 13180 KiB

Open AccessArticle

Channel-Dependent Multilayer EEG Time-Frequency Representations Combined with Transfer Learning-Based Deep CNN Framework for Few-Channel MI EEG Classification

by Ziang Liu, Kang Fan, Qin Gu and Yaduan Ruan

Bioengineering 2025, 12(6), 645; https://doi.org/10.3390/bioengineering12060645 - 12 Jun 2025

Abstract

The study of electroencephalogram (EEG) signals is crucial for understanding brain function and has extensive applications in clinical diagnosis, neuroscience, and brain–computer interface technology. This paper addresses the challenge of recognizing motor imagery EEG signals with few channels, which is essential for portable [...] Read more.

The study of electroencephalogram (EEG) signals is crucial for understanding brain function and has extensive applications in clinical diagnosis, neuroscience, and brain–computer interface technology. This paper addresses the challenge of recognizing motor imagery EEG signals with few channels, which is essential for portable and real-time applications. A novel framework is proposed that applies a continuous wavelet transform to convert time-domain EEG signals into two-dimensional time-frequency representations. These images are then concatenated into channel-dependent multilayer EEG time-frequency representations (CDML-EEG-TFR), incorporating multidimensional information of time, frequency, and channels, allowing for a more comprehensive and enriched brain representation under the constraint of few channels. By adopting a deep convolutional neural network with EfficientNet as the backbone and utilizing pre-trained weights from natural image datasets for transfer learning, the framework can simultaneously learn temporal, spatial, and channel features embedded in the CDML-EEG-TFR. Moreover, the transfer learning strategy effectively addresses the issue of data sparsity in the context of a few channels. Our approach enhances the classification accuracy of motor imagery EEG signals in few-channel scenarios. Experimental results on the BCI Competition IV 2b dataset show a significant improvement in classification accuracy, reaching 80.21%. This study highlights the potential of CDML-EEG-TFR and the EfficientNet-based transfer learning strategy in few-channel EEG signal classification, laying a foundation for practical applications and further research in medical and sports fields. Full article

(This article belongs to the Special Issue Artificial Intelligence for Biomedical Signal Processing, 2nd Edition)

► Show Figures

Graphical abstract

10 pages, 980 KiB

Open AccessBrief Report

Large-Scale Expansion of Suspension Cells in an Automated Hollow-Fiber Perfusion Bioreactor

by Eric Bräuchle, Maria Knaub, Laura Weigand, Elisabeth Ehrend, Patricia Manns, Antje Kremer, Hugo Fabre and Halvard Bonig

Bioengineering 2025, 12(6), 644; https://doi.org/10.3390/bioengineering12060644 - 12 Jun 2025

Abstract

Bioreactors enable scalable cell cultivation by providing controlled environments for temperature, oxygen, and nutrient regulation, maintaining viability and enhancing expansion efficiency. Automated systems improve reproducibility and minimize contamination risks, making them ideal for high-density cultures. While fed-batch bioreactors dominate biologics production, continuous systems [...] Read more.

Bioreactors enable scalable cell cultivation by providing controlled environments for temperature, oxygen, and nutrient regulation, maintaining viability and enhancing expansion efficiency. Automated systems improve reproducibility and minimize contamination risks, making them ideal for high-density cultures. While fed-batch bioreactors dominate biologics production, continuous systems like perfusion cultures offer superior resource efficiency and productivity. The Quantum hollow-fiber perfusion bioreactor supports cell expansion via semi-permeable capillary membranes and a closed modular design, allowing continuous media exchange while retaining key molecules. We developed a multiple-harvest protocol for suspension cells in the Quantum system, yielding 2.5 × 10¹⁰ MEL-745A cells within 29 days, with peak densities of 4 × 10⁷ cells/mL—a 15-fold increase over static cultures. Viability averaged 91.3%, with biweekly harvests yielding 3.1 × 10⁹ viable cells per harvest. Continuous media exchange required more basal media to maintain glucose and lactate levels but meaningfully less growth supplement than the 2D culture. Stable transgene expression suggested phenotypic stability. Automated processing reduced hands-on time by one-third, achieving target cell numbers 12 days earlier than 2D culture. Despite higher media use, total costs for the automated were lower compared to the manual process. Quantum enables high-density suspension cell expansion with cost advantages over conventional methods. Full article

(This article belongs to the Section Cellular and Molecular Bioengineering)

► Show Figures

Figure 1

18 pages, 11145 KiB

Open AccessArticle

Mechanistic Analysis of Fluid Dynamics and Multifactorial Impact Mechanisms in Inhaled Pharmaceutical Deposition for Chronic Respiratory Diseases

by Fuli Hu, Songhua Ma and Tianliang Hu

Bioengineering 2025, 12(6), 643; https://doi.org/10.3390/bioengineering12060643 - 12 Jun 2025

Abstract

The clinical efficacy of inhalation therapy in chronic respiratory diseases is fundamentally constrained by particle deposition patterns. This study employs computational fluid dynamics (CFD) and response surface methodology (RSM) to elucidate the mechanistic interplay of deposition determinants through multifactorial sensitivity mapping. The study [...] Read more.

The clinical efficacy of inhalation therapy in chronic respiratory diseases is fundamentally constrained by particle deposition patterns. This study employs computational fluid dynamics (CFD) and response surface methodology (RSM) to elucidate the mechanistic interplay of deposition determinants through multifactorial sensitivity mapping. The study comprises two key components: (i) the development of an accurate three-dimensional respiratory airway model spanning from the oral cavity to the fifth-generation bronchi and (ii) the integration of a Box–Behnken Design (BBD) experimental framework with computational fluid dynamics simulations. Furthermore, we developed a multifactorial regression model to analyze the synergistic interactions among deposition determinants. The study demonstrated a positive correlation between breath-holding time and drug deposition efficiency, revealing a hierarchical order of critical parameters: peak flow rate > breath-holding time > particle diameter. These findings have important implications for optimizing respiratory drug delivery strategies in clinical settings. Full article

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

► Show Figures

Figure 1

21 pages, 3230 KiB

Open AccessArticle

Active Contours Connected Component Analysis Segmentation Method of Cancerous Lesions in Unsupervised Breast Histology Images

by Vincent Majanga, Ernest Mnkandla, Zenghui Wang and Donatien Koulla Moulla

Bioengineering 2025, 12(6), 642; https://doi.org/10.3390/bioengineering12060642 - 12 Jun 2025

Abstract

Automatic segmentation of nuclei on breast cancer histology images is a basic and important step for diagnosis in a computer-aided diagnostic approach and helps pathologists discover cancer early. Nuclei segmentation remains a challenging problem due to cancer biology and the variability of tissue [...] Read more.

Automatic segmentation of nuclei on breast cancer histology images is a basic and important step for diagnosis in a computer-aided diagnostic approach and helps pathologists discover cancer early. Nuclei segmentation remains a challenging problem due to cancer biology and the variability of tissue characteristics; thus, their detection in an image is a very tedious and time-consuming task. In this context, overlapping nuclei objects present difficulties in separating them by conventional segmentation methods; thus, active contours can be employed in image segmentation. A major limitation of the active contours method is its inability to resolve image boundaries/edges of intersecting objects and segment multiple overlapping objects as a single object. Therefore, we present a hybrid active contour (connected component + active contours) method to segment cancerous lesions in unsupervised human breast histology images. Initially, this approach prepares and pre-processes data through various augmentation methods to increase the dataset size. Then, a stain normalization technique is applied to these augmented images to isolate nuclei features from tissue structures. Secondly, morphology operation techniques, namely erosion, dilation, opening, and distance transform, are used to highlight foreground and background pixels while removing overlapping regions from the highlighted nuclei objects on the image. Consequently, the connected components method groups these highlighted pixel components with similar intensity values and assigns them to their relevant labeled component to form a binary mask. Once all binary-masked groups have been determined, a deep-learning recurrent neural network (RNN) model from the Keras architecture uses this information to automatically segment nuclei objects having cancerous lesions on the image via the active contours method. This approach, therefore, uses the capabilities of connected components analysis to solve the limitations of the active contour method. This segmentation method is evaluated on an unsupervised, augmented human breast cancer histology dataset of 15,179 images. This proposed method produced a significant evaluation result of

98.71 %

accuracy score. Full article

(This article belongs to the Section Biosignal Processing)

► Show Figures

Figure 1

30 pages, 7573 KiB

Open AccessArticle

A CNN-Transformer Fusion Model for Proactive Detection of Schizophrenia Relapse from EEG Signals

by Sana Yasin, Muhammad Adeel, Umar Draz, Tariq Ali, Mohammad Hijji, Muhammad Ayaz and Ashraf M. Marei

Bioengineering 2025, 12(6), 641; https://doi.org/10.3390/bioengineering12060641 - 12 Jun 2025

Abstract

Proactively detecting schizophrenia relapse remains a critical challenge in psychiatric care, where traditional predictive models often fail to capture the complex neurophysiological and behavioral dynamics preceding recurrence. Existing methods typically rely on shallow architectures or unimodal data sources, resulting in limited sensitivity—particularly in [...] Read more.

Proactively detecting schizophrenia relapse remains a critical challenge in psychiatric care, where traditional predictive models often fail to capture the complex neurophysiological and behavioral dynamics preceding recurrence. Existing methods typically rely on shallow architectures or unimodal data sources, resulting in limited sensitivity—particularly in the early stages of relapse. In this study, we propose a CNN-Transformer fusion model that leverages the complementary strengths of Convolutional Neural Networks (CNNs) and Transformer-based architectures to process electroencephalogram (EEG) signals enriched with clinical and sentiment-derived features. This hybrid framework enables joint spatial-temporal modeling of relapse indicators, allowing for a more nuanced and patient-specific analysis. Unlike previous approaches, our model incorporates a multi-resource data fusion pipeline, improving robustness, interpretability, and clinical relevance. Experimental evaluations demonstrate a superior prediction accuracy of 97%, with notable improvements in recall and F1-score compared to leading baselines. Moreover, the model significantly reduces false negatives, a crucial factor for timely therapeutic intervention. By addressing the limitations of unimodal and superficial prediction strategies, this framework lays the groundwork for scalable, real-world applications in continuous mental health monitoring and personalized relapse prevention. Full article

(This article belongs to the Special Issue Artificial Intelligence in Biomedical Imaging, Biosignals and Healthcare)

► Show Figures

Figure 1

31 pages, 5907 KiB

Open AccessArticle

A Lightweight Breast Cancer Mass Classification Model Utilizing Simplified Swarm Optimization and Knowledge Distillation

by Wei-Chang Yeh, Wei-Chung Shia, Yun-Ting Hsu, Chun-Hui Huang and Yong-Shiuan Lee

Bioengineering 2025, 12(6), 640; https://doi.org/10.3390/bioengineering12060640 - 11 Jun 2025

Abstract

In recent years, an increasing number of women worldwide have been affected by breast cancer. Early detection is crucial, as it is the only way to identify abnormalities at an early stage. However, most deep learning models developed for classifying breast cancer abnormalities [...] Read more.

In recent years, an increasing number of women worldwide have been affected by breast cancer. Early detection is crucial, as it is the only way to identify abnormalities at an early stage. However, most deep learning models developed for classifying breast cancer abnormalities tend to be large-scale and computationally intensive, often overlooking the constraints of cost and limited computational resources. This research addresses these challenges by utilizing the CBIS-DDSM dataset and introducing a novel concatenated classification architecture and a two-stage strategy to develop an optimized, lightweight model for breast mass abnormality classification. Through data augmentation and image preprocessing, the proposed model demonstrates a superior performance compared to standalone CNN and DNN models. The two-stage strategy involves first constructing a compact model using knowledge distillation and then refining its structure with a heuristic approach known as Simplified Swarm Optimization (SSO). The experimental results confirm that knowledge distillation significantly enhances the model’s performance. Furthermore, by applying SSO’s full-variable update mechanism, the final model—SSO-Concatenated NASNetMobile (SSO-CNNM)—achieves outstanding performance metrics. It attains a compression rate of 96.17%, along with accuracy, precision, recall, and AUC scores of 96.47%, 97.4%, 94.94%, and 98.23%, respectively, outperforming other existing methods. Full article

(This article belongs to the Special Issue Artificial Intelligence in Biomedical Imaging and Biomedical Signal Processing)

► Show Figures

Figure 1

53 pages, 4286 KiB

Open AccessReview

Breast Cancer Detection Using Infrared Thermography: A Survey of Texture Analysis and Machine Learning Approaches

by Larry Ryan and Sos Agaian

Bioengineering 2025, 12(6), 639; https://doi.org/10.3390/bioengineering12060639 - 11 Jun 2025

Abstract

Breast cancer remains a leading cause of cancer-related deaths among women worldwide, highlighting the urgent need for early detection. While mammography is the gold standard, it faces cost and accessibility barriers in resource-limited areas. Infrared thermography is a promising cost-effective, non-invasive, painless, and [...] Read more.

Breast cancer remains a leading cause of cancer-related deaths among women worldwide, highlighting the urgent need for early detection. While mammography is the gold standard, it faces cost and accessibility barriers in resource-limited areas. Infrared thermography is a promising cost-effective, non-invasive, painless, and radiation-free alternative that detects tumors by measuring their thermal signatures through thermal infrared radiation. However, challenges persist, including limited clinical validation, lack of Food and Drug Administration (FDA) approval as a primary screening tool, physiological variations among individuals, differing interpretation standards, and a shortage of specialized radiologists. This survey uniquely focuses on integrating texture analysis and machine learning within infrared thermography for breast cancer detection, addressing the existing literature gaps, and noting that this approach achieves high-ranking results. It comprehensively reviews the entire processing pipeline, from image preprocessing and feature extraction to classification and performance assessment. The survey critically analyzes the current limitations, including over-reliance on limited datasets like DMR-IR. By exploring recent advancements, this work aims to reduce radiologists’ workload, enhance diagnostic accuracy, and identify key future research directions in this evolving field. Full article

(This article belongs to the Special Issue Biomedical Imaging and Data Analytics for Disease Diagnosis and Treatment, 2nd Edition)

► Show Figures

Figure 1

15 pages, 1188 KiB

Open AccessReview

Why Hoop Tension Matters: A Biomechanical Perspective on Medial Meniscus Posterior Root Tears—A Narrative Review

by Man Soo Kim, Yong In, Hyungtae Kim, Juyoung Jeong and Sueen Sohn

Bioengineering 2025, 12(6), 638; https://doi.org/10.3390/bioengineering12060638 - 11 Jun 2025

Abstract

This narrative review aims to provide an in-depth understanding of the biomechanical consequences of medial meniscus posterior root tears (MMPRTs), with a particular focus on the role of hoop tension in meniscal function. By revisiting fundamental principles such as load transmission, contact mechanics, [...] Read more.

This narrative review aims to provide an in-depth understanding of the biomechanical consequences of medial meniscus posterior root tears (MMPRTs), with a particular focus on the role of hoop tension in meniscal function. By revisiting fundamental principles such as load transmission, contact mechanics, and structural stabilization, this review elucidates how MMPRTs compromise both the integrity and function of the knee joint. The disruption of hoop tension is analyzed across various tear patterns, and through a synthesis of biomechanical experiments, the superiority and necessity of anatomical structural restoration over conservative management or meniscectomy are emphasized. A comprehensive grasp of these biomechanical foundations offers a critical perspective on the pathomechanics of MMPRTs and serves as a basis for more rational, evidence-based surgical decision-making in clinical practice. Full article

(This article belongs to the Section Biomechanics and Sports Medicine)

► Show Figures

Figure 1

13 pages, 1734 KiB

Open AccessArticle

Application of a Nomogram Model in Predicting Postoperative Delirium Following Percutaneous Coronary Intervention

by Yaxin Xiong, Ze Meng, Jiuyue Sun, Yucheng Qi, Kuo Wang, Ping Huang, Qiuyue Yang, Renliang Fan, Jiaman Guan, Mingyan Zhao and Xianglin Meng

Bioengineering 2025, 12(6), 637; https://doi.org/10.3390/bioengineering12060637 - 11 Jun 2025

Abstract

Background: Postoperative delirium is associated with an increased number of different complications, such as prolonged hospital stay, long-term cognitive impairment, and increased mortality. Therefore, early prediction of delirium after percutaneous coronary intervention (PCI) is necessary, but currently, there is still a lack [...] Read more.

Background: Postoperative delirium is associated with an increased number of different complications, such as prolonged hospital stay, long-term cognitive impairment, and increased mortality. Therefore, early prediction of delirium after percutaneous coronary intervention (PCI) is necessary, but currently, there is still a lack of reliable and effective prediction models for such patients. Methods: All data used in this study were derived from the MIMIC-IV database. Multivariable Cox regression was employed to analyze the data, and the performance of the newly developed nomogram was evaluated based on the area under the receiver operating characteristic curve (AUC). The clinical value of the prediction model was tested using decision curve analysis (DCA). Results: A total of 313 PCI patients in the intensive care unit (ICU) were included in the analysis, comprising 219 in the training cohort and 94 in the testing cohort. Twenty variables were selected for model development. Multivariable Cox regression revealed that benzodiazepine use, vasoactive drug therapy, age, white blood cell count (WBC), and serum potassium were independent risk factors for predicting the occurrence of delirium after PCI. The AUC values for predicting delirium occurrence in the training and validation cohorts were 0.771 and 0.743, respectively. Conclusions: This study has identified several important demographic and laboratory parameters associated with the occurrence of delirium after PCI, and used them to establish a more accurate and convenient nomogram model to predict the occurrence of postoperative delirium in such patients. Full article

(This article belongs to the Special Issue Biomedical Application of Big Data and Artificial Intelligence—Second Edition)

► Show Figures

Figure 1

35 pages, 4377 KiB

Open AccessArticle

Liver Semantic Segmentation Method Based on Multi-Channel Feature Extraction and Cross Fusion

by Chenghao Zhang, Lingfei Wang, Chunyu Zhang, Yu Zhang, Peng Wang and Jin Li

Bioengineering 2025, 12(6), 636; https://doi.org/10.3390/bioengineering12060636 - 11 Jun 2025

Abstract

Semantic segmentation plays a critical role in medical image analysis, offering indispensable information for the diagnosis and treatment planning of liver diseases. However, due to the complex anatomical structure of the liver and significant inter-patient variability, the current methods exhibit notable limitations in [...] Read more.

Semantic segmentation plays a critical role in medical image analysis, offering indispensable information for the diagnosis and treatment planning of liver diseases. However, due to the complex anatomical structure of the liver and significant inter-patient variability, the current methods exhibit notable limitations in feature extraction and fusion, which pose a major challenge to achieving accurate liver segmentation. To address these challenges, this study proposes an improved U-Net-based liver semantic segmentation method that enhances segmentation performance through optimized feature extraction and fusion mechanisms. Firstly, a multi-scale input strategy is employed to account for the variability in liver features at different scales. A multi-scale convolutional attention (MSCA) mechanism is integrated into the encoder to aggregate multi-scale information and improve feature representation. Secondly, an atrous spatial pyramid pooling (ASPP) module is incorporated into the bottleneck layer to capture features at various receptive fields using dilated convolutions, while global pooling is applied to enhance the acquisition of contextual information and ensure efficient feature transmission. Furthermore, a Channel Transformer module replaces the traditional skip connections to strengthen the interaction and fusion between encoder and decoder features, thereby reducing the semantic gap. The effectiveness of this method was validated on integrated public datasets, achieving an Intersection over Union (IoU) of 0.9315 for liver segmentation tasks, outperforming other mainstream approaches. This provides a novel solution for precise liver image segmentation and holds significant clinical value for liver disease diagnosis and treatment. Full article

(This article belongs to the Special Issue Machine Learning and Deep Learning Applications in Healthcare)

► Show Figures

Figure 1

19 pages, 3300 KiB

Open AccessArticle

Electric Potential of Chlorella sp. Microalgae Biomass in Microbial Fuel Cells (MFCs)

by Rickelmi Agüero-Quiñones, Magaly De La Cruz-Noriega and Walter Rojas-Villacorta

Bioengineering 2025, 12(6), 635; https://doi.org/10.3390/bioengineering12060635 - 11 Jun 2025

Abstract

The projected global energy demand for 2050 drives the imperative search for alternative and environmentally friendly energy sources. An emerging and promising alternative is microbial fuel cells assisted with microalgae. This research evaluated the potential of Chlorella sp. biomass in electricity production using [...] Read more.

The projected global energy demand for 2050 drives the imperative search for alternative and environmentally friendly energy sources. An emerging and promising alternative is microbial fuel cells assisted with microalgae. This research evaluated the potential of Chlorella sp. biomass in electricity production using microbial fuel cells (MFCs) with a single chamber and activated carbon and zinc electrodes at the laboratory scale over 20 days of operation. Maximum values of voltage (1271 ± 2.52 mV), current (4.77 ± 0.02 mA), power density (247.514 mW/cm²), current density (0.551 mA/cm²), and internal resistance (200.83 ± 0.327 Ω) were obtained. The biomass-maintained pH values of 7.32 ± 0.03–7.74 ± 0.02 and peaks of electrical conductivity of 2450 ± 17.1 µS/cm and oxidation-reduction potential of 952 ± 20 mV were reached. Meanwhile, cell density and absorbance increased to average values of 2.2933 × 10⁷ ± 1.15 × 10⁶ cells/mL and 3.471 ± 0.195 absorbance units (AU), respectively. Scanning electron microscopy micrographs allowed the observation of filamentous structures of the formed biofilm attached to carbon particles, and energy-dispersive X-ray spectroscopy spectra of the anodes determined the predominance of oxygen, carbon, silicon, aluminum, and iron. Finally, this research demonstrates the great potential of Chlorella sp. biomass for sustainable bioelectricity generation in MFCs. Full article

(This article belongs to the Special Issue Development of Bioelectrochemical Systems for Environmental Engineering)

► Show Figures

Figure 1

13 pages, 1548 KiB

Open AccessArticle

Application of the Brillouin Optical Scanning System in the Regional Corneal Biomechanical Evaluation of Keratoconus and Its Correlation with Corvis ST Parameters

by Qiuruo Jiang, Yichen Sun, Zhanhao Gu, Lumeng Wang, Yiqiang Wu, Jialu Chen, Zhiyi Chen, Xiaobo Zheng and Shihao Chen

Bioengineering 2025, 12(6), 634; https://doi.org/10.3390/bioengineering12060634 - 11 Jun 2025

Abstract

(1) Background: The early diagnosis of keratoconus is critical for prognosis. Traditional methods like ORA and Corvis ST measure overall corneal biomechanics but lack regional specificity and are affected by intraocular pressure. In contrast, Brillouin microscopy assesses regional corneal biomechanics without such limitations; [...] Read more.

(1) Background: The early diagnosis of keratoconus is critical for prognosis. Traditional methods like ORA and Corvis ST measure overall corneal biomechanics but lack regional specificity and are affected by intraocular pressure. In contrast, Brillouin microscopy assesses regional corneal biomechanics without such limitations; (2) Methods: In total, 25 keratoconus patients and 28 healthy controls were included in this study. Corneal biomechanics were measured using the BOSS system (Brillouin Optical Scanning System) in a 10-point mode within an 8 mm diameter, and included the mean, maximum, minimum and standard Brillouin shift. The Corvis ST parameters extracted included the CBI (Corneal Biomechanical Index), CCBI (Corvis Biomechanical Index for Chinese populations), SSI (Stress–Strain Index), DA (Deformation Amplitude), IIR (Inverse Integrated Radius), and SP-A1 (Stiffness Parameter at First Applanation); (3) Results: BOSS showed significant differences in the inferior nasal region (p = 0.004) and central region (p = 0.029) between groups, but not in peripheral regions (p = 0.781). In a comparison of the Brillouin frequency shifts measured between groups, there was no difference in the Mean (p = 0.452) and Max (p = 0.487), but the Min (p = 0.003), Standard (p = 0.000), and Max–Min (p = 0.006) all showed differences. Corvis ST identified significant differences in six parameters (CBI, CCBI, SSI, DA, IIR, and SP-A1) between groups (p < 0.001). Correlations were found between the BOSS and Corvis ST results, with moderate correlations in the inferior nasal region; (4) Conclusions: The BOSS Brillouin microscope can provide an accurate diagnostic evaluation for the corneal biomechanical differences between normal eyes and keratoconus, independent of IOP (Intraocular Pressure) and CCT (Central Corneal Thickness), with a good correlation with Corvis ST, especially in assessing regional biomechanics. Full article

► Show Figures

Graphical abstract

12 pages, 505 KiB

Open AccessArticle

Assessment of Possibility of Using Ultrasound Imaging in Treatment of Stress Urinary Incontinence in Women—Preliminary Study

by Gabriela Kołodyńska, Maciej Zalewski, Aleksandra Piątek, Anna Mucha, Krystyna Rożek-Piechura and Waldemar Andrzejewski

Bioengineering 2025, 12(6), 633; https://doi.org/10.3390/bioengineering12060633 - 10 Jun 2025

Abstract

The number of people suffering from urinary incontinence increases every year. Along this trend, the knowledge of society increases regarding the various methods available for treating this ailment. Both patients and researchers are constantly looking for new treatments for urinary incontinence. One of [...] Read more.

The number of people suffering from urinary incontinence increases every year. Along this trend, the knowledge of society increases regarding the various methods available for treating this ailment. Both patients and researchers are constantly looking for new treatments for urinary incontinence. One of the new solutions is sonofeedback of the pelvic floor muscles, which may help to strengthen them and thus reduce the problem. The aim of this study was to evaluate the effectiveness of sonofeedback and transvaginal electrostimulation in increasing the bioelectrical activity of pelvic floor muscles in postmenopausal women with stress urinary incontinence. Sixty women with stress urinary incontinence were enrolled in the study. The patients were divided into two groups: A, where sonofeedback was used, and B, where electrostimulation of the pelvic floor muscles was performed with biofeedback training. In patients, the resting bioelectrical activity of the pelvic floor muscles was assessed using an electromyograph. The assessment of the resting bioelectrical activity of the pelvic floor muscles was performed before the therapy, after the 5th training, and after the therapy. It was observed that after the end of the therapy, the average bioelectrical potential increased by 1.1 µV compared with the baseline in group A. It can be suggested that the sonofeedback method is comparatively effective in reducing symptoms that are associated with urinary incontinence as an electrostimulation method with biofeedback training. Full article

(This article belongs to the Section Biosignal Processing)

► 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