Submit to Bioengineering Review for Bioengineering Propose a Special Issue

Journal Menu

Journal Browser

► Journal Browser

Machine Learning and Artificial Intelligence for Biomedical Applications, 3rd Edition

Special Issue Editors
Special Issue Information
Keywords
Benefits of Publishing in a Special Issue
Related Special Issues
Published Papers

A special issue of Bioengineering (ISSN 2306-5354). This special issue belongs to the section "Biosignal Processing".

Deadline for manuscript submissions: closed (31 December 2025) | Viewed by 15832

Share This Special Issue

Special Issue Editor

Dr. Crescenzio Gallo

E-Mail Website
Guest Editor

Department of Clinical and Experimental Medicine, Università degli Studi di Foggia, 71122 Foggia, FG, Italy
Interests: bioinformatics; artificial intelligence
Special Issues, Collections and Topics in MDPI journals

Special Issue Information

Dear Colleagues,

In recent years, an increase in the accuracy of information technology has led to several scientific breakthroughs. The first researchers to benefit from improved hardware components have been the developers of artificial intelligence algorithms, who have been able to apply these algorithms in several scientific fields, including biomedicine. Biomedicine is a field of medicine that applies the principles of biology and natural sciences to the development of relevant technologies for healthcare. The combination of artificial intelligence algorithms and biomedicine has led to many applications, such as image analysis of human organs using magnetic resonance images (MRI); DNA/RNA sequencing and protein structure interactions and predictions; and analysis of different biosignals via methods involving electroencephalograms (EEGs), electromyography (EMGs), and electrocardiograms (ECGs).

In this context, machine learning algorithms enable us to learn from observational data and construct highly accurate artificial intelligence models to support the physician. However, obtaining models with high accuracy may not be enough, as AI-based biomedical decisions must be understandable to the physician. Therefore, it is necessary to equip machine learning methods with explainability capacity, leading to explainable artificial intelligence techniques that enable the physician to understand the decisions suggested by the models they use.

This is the third volume of our Special Issue series "Machine Learning and Artificial Intelligence for Biomedical Applications". Please feel free to download and read the first two volumes via the following links:
https://www.mdpi.com/journal/bioengineering/special_issues/39708P1H4A
https://www.mdpi.com/journal/bioengineering/special_issues/483SCWZ885

Dr. Crescenzio Gallo
Guest Editor

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 250 words) can be sent to the Editorial Office for assessment.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Bioengineering is an international peer-reviewed open access monthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2700 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Keywords

artificial intelligence models
biomedicine
machine learning methods
artificial neural networks
precision medicine
personalized health care

Benefits of Publishing in a Special Issue

Ease of navigation: Grouping papers by topic helps scholars navigate broad scope journals more efficiently.
Greater discoverability: Special Issues support the reach and impact of scientific research. Articles in Special Issues are more discoverable and cited more frequently.
Expansion of research network: Special Issues facilitate connections among authors, fostering scientific collaborations.
External promotion: Articles in Special Issues are often promoted through the journal's social media, increasing their visibility.
Reprint: MDPI Books provides the opportunity to republish successful Special Issues in book format, both online and in print.

Further information on MDPI's Special Issue policies can be found here.

Related Special Issues

Published Papers (10 papers)

Download All Papers

Order results

Result details

Show export options Show export options

Select all

Export citation of selected articles as:

Research

Jump to: Review, Other

15 pages, 1579 KB

Open AccessArticle

Digital Twin and Artificial Intelligence Technologies to Assess the Type IA Endoleak

by Sungsin Cho, Hyangkyoung Kim and Jinhyun Joh

Bioengineering 2026, 13(1), 1; https://doi.org/10.3390/bioengineering13010001 - 19 Dec 2025

Viewed by 419

Abstract

Background/Objectives: Endovascular aneurysm repair (EVAR) is the standard treatment for abdominal aortic aneurysms, but the risk of endoleak compromises its effectiveness. Type IA endoleak, stemming from an inadequate proximal seal, is the most critical complication associated with the highest risk of rupture. Current preoperative planning relies on static anatomical measurements from computed tomography angiography that fail to predict seal failure due to dynamic biomechanical forces. This study aimed to retrospectively validate the predictive accuracy of a novel physics-informed digital twin and artificial intelligence (AI) model for predicting type IA endoleak risk compared to conventional static planning methods. Methods: This was a retrospective, single-center proof-of-concept validation study involving 15 patients who underwent elective EVAR (5 with confirmed type IA endoleak and 10 without type IA endoleak). A patient-specific digital twin was created for each case to simulate stent-graft deployment and capture the dynamic biomechanical interaction with the aortic wall. A logistic regression AI model processed over 16,000 biomechanical measurements to generate a single, objective metric of the endoleak risk index (ERI). The predictive performance of the ERI (using a cutoff of 0.80) was assessed and compared against a 1:3 propensity score-matched conventional control group (n = 45) who received traditional anatomical-based planning. Results: The mean ERI was significantly higher in the endoleak-positive group (0.85 ± 0.10) compared to the endoleak-negative group (0.39 ± 0.11) (p = 0.011). The digital twin/AI model demonstrated superior predictive capability, achieving an overall accuracy of 80% (95% CI: 51.9–95.7) and an area under the curve (AUC) of 0.85 (95% CI: 0.58–0.99). Crucially, the model achieved a sensitivity of 100% and a negative predictive value (NPV) of 100%, correctly identifying all high-risk cases and ruling out endoleak in all low-risk cases. In stark contrast, the matched conventional planning group achieved an overall accuracy of only 51.1% and an AUC of 0.54. Conclusion: This physics-informed digital twin and AI framework successfully validated its capability to accurately and objectively predict the risk of type IA endoleak following EVAR. The derived ERI offers a significant quantitative advantage over traditional static anatomical measurements, establishing it as a highly reliable safety tool (100% NPV) for ruling out endoleak risk. This technology represents a critical advancement toward personalized EVAR planning, enabling surgeons to proactively identify high-risk anatomies and adjust treatment strategies to minimize post-procedural complications. Further large-scale, multicenter prospective trials are necessary to confirm these findings and support clinical adoption. Full article

(This article belongs to the Special Issue Machine Learning and Artificial Intelligence for Biomedical Applications, 3rd Edition)

► Show Figures

Figure 1

27 pages, 11177 KB

Open AccessArticle

Robust Segmentation of Lung Proton and Hyperpolarized Gas MRI with Vision Transformers and CNNs: A Comparative Analysis of Performance Under Artificial Noise

by Ramtin Babaeipour, Matthew S. Fox, Grace Parraga and Alexei Ouriadov

Bioengineering 2025, 12(8), 808; https://doi.org/10.3390/bioengineering12080808 - 28 Jul 2025

Cited by 1 | Viewed by 1050

Abstract

Accurate segmentation in medical imaging is essential for disease diagnosis and monitoring, particularly in lung imaging using proton and hyperpolarized gas MRI. However, image degradation due to noise and artifacts—especially in hyperpolarized gas MRI, where scans are acquired during breath-holds—poses challenges for conventional segmentation algorithms. This study evaluates the robustness of deep learning segmentation models under varying Gaussian noise levels, comparing traditional convolutional neural networks (CNNs) with modern Vision Transformer (ViT)-based models. Using a dataset of proton and hyperpolarized gas MRI slices from 56 participants, we trained and tested Feature Pyramid Network (FPN) and U-Net architectures with both CNN (VGG16, VGG19, ResNet152) and ViT (MiT-B0, B3, B5) backbones. Results showed that ViT-based models, particularly those using the SegFormer backbone, consistently outperformed CNN-based counterparts across all metrics and noise levels. The performance gap was especially pronounced in high-noise conditions, where transformer models retained higher Dice scores and lower boundary errors. These findings highlight the potential of ViT-based architectures for deployment in clinically realistic, low-SNR environments such as hyperpolarized gas MRI, where segmentation reliability is critical. Full article

(This article belongs to the Special Issue Machine Learning and Artificial Intelligence for Biomedical Applications, 3rd Edition)

► Show Figures

Figure 1

18 pages, 24638 KB

Open AccessArticle

Accelerating Wound Healing Through Deep Reinforcement Learning: A Data-Driven Approach to Optimal Treatment

by Fan Lu, Ksenia Zlobina, Prabhat Baniya, Houpu Li, Nicholas Rondoni, Narges Asefifeyzabadi, Wan Shen Hee, Maryam Tebyani, Kaelan Schorger, Celeste Franco, Michelle Bagood, Mircea Teodorescu, Marco Rolandi, Rivkah Isseroff and Marcella Gomez

Bioengineering 2025, 12(7), 756; https://doi.org/10.3390/bioengineering12070756 - 11 Jul 2025

Cited by 2 | Viewed by 1377

Abstract

Advancements in bioelectronic sensors and actuators have paved the way for real-time monitoring and control of the progression of wound healing. Real-time monitoring allows for precise adjustment of treatment strategies to align them with an individual’s unique biological response. However, due to the complexities of human–drug interactions and a lack of predictive models, it is challenging to determine how one should adjust drug dosage to achieve the desired biological response. This work proposes an adaptive closed-loop control framework that integrates deep learning, optimal control, and reinforcement learning to update treatment strategies in real time, with the goal of accelerating wound closure. The proposed approach eliminates the need for mathematical modeling of complex nonlinear wound-healing dynamics. We demonstrate the convergence of the controller via an in silico experimental setup, where the proposed approach successfully accelerated the wound-healing process by 17.71%. Finally, we share the experimental setup and results of an in vivo implementation to highlight the translational potential of our work. Our data-driven model suggests that the treatment strategy, as determined by our deep reinforcement learning algorithm, results in an accelerated onset of inflammation and subsequent transition to proliferation in a porcine wound model. Full article

(This article belongs to the Special Issue Machine Learning and Artificial Intelligence for Biomedical Applications, 3rd Edition)

► Show Figures

Figure 1

22 pages, 4293 KB

Open AccessArticle

Speech-Based Parkinson’s Detection Using Pre-Trained Self-Supervised Automatic Speech Recognition (ASR) Models and Supervised Contrastive Learning

by Hadi Sedigh Malekroodi, Nuwan Madusanka, Byeong-il Lee and Myunggi Yi

Bioengineering 2025, 12(7), 728; https://doi.org/10.3390/bioengineering12070728 - 1 Jul 2025

Cited by 4 | Viewed by 3541

Abstract

Diagnosing Parkinson’s disease (PD) through speech analysis is a promising area of research, as speech impairments are often one of the early signs of the disease. This study investigates the efficacy of fine-tuning pre-trained Automatic Speech Recognition (ASR) models, specifically Wav2Vec 2.0 and HuBERT, for PD detection using transfer learning. These models, pre-trained on large unlabeled datasets, can be capable of learning rich speech representations that capture acoustic markers of PD. The study also proposes the integration of a supervised contrastive (SupCon) learning approach to enhance the models’ ability to distinguish PD-specific features. Additionally, the proposed ASR-based features were compared against two common acoustic feature sets: mel-frequency cepstral coefficients (MFCCs) and the extended Geneva minimalistic acoustic parameter set (eGeMAPS) as a baseline. We also employed a gradient-based method, Grad-CAM, to visualize important speech regions contributing to the models’ predictions. The experiments, conducted using the NeuroVoz dataset, demonstrated that features extracted from the pre-trained ASR models exhibited superior performance compared to the baseline features. The results also reveal that the method integrating SupCon consistently outperforms traditional cross-entropy (CE)-based models. Wav2Vec 2.0 and HuBERT with SupCon achieved the highest F1 scores of 90.0% and 88.99%, respectively. Additionally, their AUC scores in the ROC analysis surpassed those of the CE models, which had comparatively lower AUCs, ranging from 0.84 to 0.89. These results highlight the potential of ASR-based models as scalable, non-invasive tools for diagnosing and monitoring PD, offering a promising avenue for the early detection and management of this debilitating condition. Full article

(This article belongs to the Special Issue Machine Learning and Artificial Intelligence for Biomedical Applications, 3rd Edition)

► Show Figures

Figure 1

19 pages, 2124 KB

Open AccessArticle

A Unified Deep Learning Ensemble Framework for Voice-Based Parkinson’s Disease Detection and Motor Severity Prediction

by Madjda Khedimi, Tao Zhang, Chaima Dehmani, Xin Zhao and Yanzhang Geng

Bioengineering 2025, 12(7), 699; https://doi.org/10.3390/bioengineering12070699 - 27 Jun 2025

Cited by 1 | Viewed by 2643

Abstract

This study presents a hybrid ensemble learning framework for the joint detection and motor severity prediction of Parkinson’s disease (PD) using biomedical voice features. The proposed architecture integrates a deep multimodal fusion model with dense expert pathways, multi-head self-attention, and multitask output branches to simultaneously perform binary classification and regression. To ensure data quality and improve model generalization, preprocessing steps included outlier removal via Isolation Forest, two-stage feature scaling (RobustScaler followed by MinMaxScaler), and augmentation through polynomial and interaction terms. Borderline-SMOTE was employed to address class imbalance in the classification task. To enhance prediction performance, ensemble learning strategies were applied by stacking outputs from the fusion model with tree-based regressors (Random Forest, Gradient Boosting, and XGBoost), using diverse meta-learners including XGBoost, Ridge Regression, and a deep neural network. Among these, the Stacking Ensemble with XGBoost (SE-XGB) achieved the best results, with an R² of 99.78% and RMSE of 0.3802 for UPDRS regression and 99.37% accuracy for PD classification. Comparative analysis with recent literature highlights the superior performance of our framework, particularly in regression settings. These findings demonstrate the effectiveness of combining advanced feature engineering, deep learning, and ensemble meta-modeling for building accurate and generalizable models in voice-based PD monitoring. This work provides a scalable foundation for future clinical decision support systems. Full article

(This article belongs to the Special Issue Machine Learning and Artificial Intelligence for Biomedical Applications, 3rd Edition)

► Show Figures

Figure 1

11 pages, 1016 KB

Open AccessArticle

Graph Representation Learning for the Prediction of Medication Usage in the UK Biobank Based on Pharmacogenetic Variants

by Bill Qi and Yannis J. Trakadis

Bioengineering 2025, 12(6), 595; https://doi.org/10.3390/bioengineering12060595 - 31 May 2025

Viewed by 1761

Abstract

Ineffective treatment and side effects are associated with high burdens for the patient and society. We investigated the application of graph representation learning (GRL) for predicting medication usage based on individual genetic data in the United Kingdom Biobank (UKBB). A graph convolutional network (GCN) was used to integrate interconnected biomedical entities in the form of a knowledge graph as part of a machine learning (ML) prediction model. Data from The Pharmacogenomics Knowledgebase (PharmGKB) was used to construct a biomedical knowledge graph. Individual genetic data (n = 485,754) from the UKBB was obtained and preprocessed to match with pharmacogenetic variants in the PharmGKB. Self-reported medication usage labels were obtained from UKBB data field 20003. We hypothesize that pharmacogenetic variants can predict the impact of medications on individuals. We assume that an individual using a medication on a regular basis experiences a net benefit (vs. side-effects) from the medication. ML models were trained to predict medication usage for 264 medications. The GCN model significantly outperformed both a baseline logistic regression model (p-value: 1.53 × 10⁻⁹) and a deep neural network model (p-value: 8.68 × 10⁻⁸). The GCN model also significantly outperformed a GCN model trained using a random graph (GCN-random) (p-value: 5.44 × 10⁻⁹). A consistent trend of medications with higher sample sizes having better performance was observed, and for several medications, a high relative rank of the medication (among multiple medications) was associated with greater than 2-fold higher odds of usage of the medication. In conclusion, a graph-based ML approach could be useful in advancing precision medicine by prioritizing medications that a patient may need based on their genetic data. However, further research is needed to improve the quality and quantity of genetic data and to validate our approach using more reliable medication labels. Full article

(This article belongs to the Special Issue Machine Learning and Artificial Intelligence for Biomedical Applications, 3rd Edition)

► Show Figures

Graphical abstract

16 pages, 6739 KB

Open AccessArticle

A Multitask Deep Learning Model for Predicting Myocardial Infarction Complications

by Fazliddin Makhmudov, Normakhmad Ravshanov, Dilshot Akhmedov, Oleg Pekos, Dilmurod Turimov and Young-Im Cho

Bioengineering 2025, 12(5), 520; https://doi.org/10.3390/bioengineering12050520 - 14 May 2025

Cited by 2 | Viewed by 2409

Abstract

Myocardial infarction is one of the most severe forms of ischemic heart disease, associated with high mortality and disability worldwide. The accurate and reliable prediction of adverse cardiovascular events is critical for developing effective treatment strategies and improving outcomes in cardiac rehabilitation. Traditional prognostic models, such as the GRACE and TIMI scores, often lack the flexibility to incorporate a wide range of contemporary clinical predictors. Therefore, machine learning methods, particularly deep neural networks, have recently emerged as promising alternatives capable of enhancing predictive accuracy and enabling more personalized care. This study presents a multitask deep learning model designed to simultaneously address two related tasks: multidimensional binary classification of myocardial infarction complications and multiclass classification of mortality causes. The model was trained on a dataset of 1700 patients, encompassing 111 clinical and demographic features. Experimental results demonstrate high predictive accuracy and the model’s capacity to capture complex interactions among risk factors, suggesting its potential as a valuable tool for clinical decision support in cardiology. Comparative analysis confirms that the proposed multitask approach performs comparably to, or better than, conventional machine learning models. Future research will focus on refining the model and validating its generalizability in real-world clinical environments. Full article

(This article belongs to the Special Issue Machine Learning and Artificial Intelligence for Biomedical Applications, 3rd Edition)

► Show Figures

Figure 1

14 pages, 1138 KB

Open AccessArticle

Utility of Clustering in Mortality Risk Stratification in Pulmonary Hypertension

by Pasquale Tondo, Lucia Tricarico, Giuseppe Galgano, Maria Pia C. Varlese, Daphne Aruanno, Crescenzio Gallo, Giulia Scioscia, Natale D. Brunetti, Michele Correale and Donato Lacedonia

Bioengineering 2025, 12(4), 408; https://doi.org/10.3390/bioengineering12040408 - 11 Apr 2025

Cited by 1 | Viewed by 811

Abstract

Background: Pulmonary hypertension (PH) is a condition characterized by increased pressure in the pulmonary arteries with poor prognosis and, therefore, an optimal management is necessary. The study’s aim was to search for PH phenotypes and develop a predictive model of five-year mortality using machine learning (ML) algorithms. Methods: This multicenter study was conducted on 122 PH patients. Clinical and demographic data were collected and then used to identify phenotypes through clustering. Subsequently, a predictive model was performed by different ML algorithms. Results: Three PH clusters were identified: Cluster 1 (mean age 68.57 ± 10.54) includes 57% females, 69% from non-respiratory PH groups, and better cardiac (NYHA class 2.61 ± 0.84) and respiratory function (FEV1% 78.78 ± 21.54); Cluster 2 includes 50% females, mean age of 71.36 ± 8.32 years, 44% from PH group 3, worse respiratory function (FEV 1% 68.12 ± 10.20); intermediate cardiac function (NYHA class 3.18 ± 0.49) and significantly higher mortality (75%); Cluster 3 represents the youngest cluster (mean age 61.11 ± 13.50) with 65% males, 81% from non-respiratory PH groups, intermediate respiratory function (FEV1% 70.51 ± 17.91) and worse cardiac performance (NYHA class 3.22 ± 0.58). After testing ML models, logistic regression showed the best predictive performance (AUC = 0.835 and accuracy = 0.744) and identified three mortality-risk factors: age, NYHA class, and number of medications taken. Conclusions: The results suggest that the integration of ML into clinical practice can improve risk stratification to optimize treatment strategies and improve outcomes for PH patients. Full article

(This article belongs to the Special Issue Machine Learning and Artificial Intelligence for Biomedical Applications, 3rd Edition)

► Show Figures

Figure 1

Review

Jump to: Research, Other

22 pages, 462 KB

Open AccessReview

Artificial Intelligence in Tetralogy of Fallot: From Prenatal Diagnosis to Lifelong Management: A Narrative Review

by Tiziana Fragasso, Davide Passaro, Alessandra Toscano, Antonio Amodeo, Alberto Eugenio Tozzi and Giorgia Grutter

Bioengineering 2025, 12(12), 1349; https://doi.org/10.3390/bioengineering12121349 - 10 Dec 2025

Viewed by 691

Abstract

Artificial intelligence (AI) is rapidly transforming cardiovascular medicine, with profound implications for congenital heart disease (CHD). Tetralogy of Fallot (ToF), the most common cyanotic disease, requires lifelong surveillance and complex management because of late complications such as pulmonary regurgitation, arrhythmias, and right ventricular dysfunction. This review synthesizes current evidence on AI applications across the continuum of ToF care—from prenatal diagnosis to adulthood follow-up. We examine advances in imaging, perioperative planning, intraoperative monitoring, intensive care, and long-term surveillance, including wearable and implantable technologies. Machine learning (ML), deep learning (DL), and natural language processing (NLP) are revolutionizing diagnostic accuracy, risk stratification, surgical decision-making, and personalized long-term care. The future lies in the integration of multimodal data, including imaging, electronic health records (EHRs), genomic information, and continuous monitoring, to support precision medicine. Challenges remain regarding dataset limitations, interpretability, regulatory standards, and ethical concerns. Nevertheless, ongoing innovation and collaboration between clinicians, engineers, and regulators promise a new era in congenital cardiology. By embedding AI throughout the patient journey, healthcare systems may improve outcomes and quality of life for individuals with ToF. Full article

(This article belongs to the Special Issue Machine Learning and Artificial Intelligence for Biomedical Applications, 3rd Edition)

► Show Figures

Figure 1

Other

Jump to: Research, Review

42 pages, 845 KB

Open AccessSystematic Review

A Systematic Review of Contrastive Learning in Medical AI: Foundations, Biomedical Modalities, and Future Directions

by George Obaido, Ibomoiye Domor Mienye, Kehinde Aruleba, Chidozie Williams Chukwu, Ebenezer Esenogho and Cameron Modisane

Bioengineering 2026, 13(2), 176; https://doi.org/10.3390/bioengineering13020176 - 2 Feb 2026

Abstract

Medical artificial intelligence (AI) systems depend heavily on high-quality data representations to support accurate prediction, diagnosis, and clinical decision-making. However, the availability of large, well-annotated medical datasets is often constrained by cost, privacy concerns, and the need for expert labeling, motivating growing interest in self-supervised representation learning. Among these approaches, contrastive learning has emerged as one of the most influential paradigms, driving major advances in representation learning across computer vision and natural language processing. This paper presents a comprehensive review of contrastive learning in medical AI, highlighting its theoretical foundations, methodological developments, and practical applications in medical imaging, electronic health records, physiological signal analysis, and genomics. Furthermore, we identify recurring challenges, including pair construction, sensitivity to data augmentations, and inconsistencies in evaluation protocols, while discussing emerging trends such as multimodal alignment, federated learning, and privacy-preserving frameworks. Through a synthesis of current developments and open research directions, this review provides insights to advance data-efficient, reliable, and generalizable medical AI systems. Full article

(This article belongs to the Special Issue Machine Learning and Artificial Intelligence for Biomedical Applications, 3rd Edition)

Show export options Show export options

Select all

Export citation of selected articles as:

Displaying articles 1-10

Journal Menu

Journal Browser

Machine Learning and Artificial Intelligence for Biomedical Applications, 3rd Edition

Share This Special Issue

Special Issue Editor

Special Issue Information

Keywords

Benefits of Publishing in a Special Issue

Related Special Issues

Published Papers (10 papers)

Research

Review

Other

Further Information

Guidelines

MDPI Initiatives

Follow MDPI