Applied Sciences

Research

Jump to: Review, Other

22 pages, 2918 KB

Open AccessArticle

MV-RiskNet: Multi-View Attention-Based Deep Learning Model for Regional Epidemic Risk Prediction and Mapping

by Beyzanur Okudan and Abdullah Ammar Karcioglu

Appl. Sci. 2026, 16(4), 2135; https://doi.org/10.3390/app16042135 - 22 Feb 2026

Viewed by 420

Abstract

Regional epidemic risk prediction requires holistic modeling of heterogeneous data sources such as demographic structure, health capacity, geographical features and human mobility. In this study, a unique and multi-modal epidemiological data set integrating demographic, health, geographic and mobility indicators of Türkiye and its [...] Read more.

Regional epidemic risk prediction requires holistic modeling of heterogeneous data sources such as demographic structure, health capacity, geographical features and human mobility. In this study, a unique and multi-modal epidemiological data set integrating demographic, health, geographic and mobility indicators of Türkiye and its neighboring countries was collected. Türkiye’s neighboring countries are Greece, Bulgaria, Georgia, Armenia, Iran, and Iraq. This dataset, created by combining raw data from these neighboring countries, provides a comprehensive regional representation that allows for both quantitative classification and spatial mapping of epidemiological risk. To address the class imbalance problem, Conditional GAN (CGAN), a class-conditional synthetic example generation approach that enhances high-risk category representation was used. In this study, we proposed a multi-view deep learning model named MV-RiskNet, which effectively models the multi-dimensional data structure by processing each view into independent subnetworks and integrating the representations with an attention-based fusion mechanism for regional epidemic risk prediction. Experimental studies were compared using Multi-Layer Perceptron (MLP), Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), Autoencoder classifier, and Graph Convolutional Network (GCN) models. The proposed MV-RiskNet with CGAN model achieved better results compared to other models, with 97.22% accuracy and 97.40% F1-score. The generated risk maps reveal regional clustering patterns in a spatially consistent manner, while attention analyses show that demographic and geographic features are the dominant determinants, while mobility plays a complementary role, especially in high-risk regions. Full article

(This article belongs to the Special Issue Artificial Intelligence Applications in Healthcare and Precision Medicine, 2nd Edition)

► Show Figures

Graphical abstract

21 pages, 2013 KB

Open AccessArticle

Machine Learning Models for Reliable Gait Phase Detection Using Lower-Limb Wearable Sensor Data

by Muhammad Fiaz, Rosita Guido and Domenico Conforti

Appl. Sci. 2026, 16(3), 1397; https://doi.org/10.3390/app16031397 - 29 Jan 2026

Viewed by 674

Abstract

Accurate gait-phase detection is essential for rehabilitation monitoring, prosthetic control, and human–robot interaction. Artificial intelligence supports continuous, personalized mobility assessment by extracting clinically meaningful patterns from wearable sensors. A richer view of gait dynamics can be achieved by integrating additional signals, including inertial, [...] Read more.

Accurate gait-phase detection is essential for rehabilitation monitoring, prosthetic control, and human–robot interaction. Artificial intelligence supports continuous, personalized mobility assessment by extracting clinically meaningful patterns from wearable sensors. A richer view of gait dynamics can be achieved by integrating additional signals, including inertial, plantar flex, footswitch, and EMG data, leading to more accurate and informative gait analysis. Motivated by these needs, this study investigates discrete gait-phase recognition for the right leg using a multi-subject IMU dataset collected from lower-limb sensors. IMU recordings were segmented into 128-sample windows across 23 channels, and each window was flattened into a 2944-dimensional feature vector. To ensure reliable ground-truth labels, we developed an automatic relabeling pipeline incorporating heel-strike and toe-off detection, adaptive threshold tuning, and sensor fusion across sensor modalities. These windowed vectors were then used to train a comprehensive suite of machine learning models, including Random Forests, Extra Trees, k-Nearest Neighbors, XGBoost, and LightGBM. All models underwent systematic hyperparameter tuning, and their performance was assessed through k-fold cross-validation. The results demonstrate that tree-based ensemble models provide accurate and stable gait-phase classification with accuracy exceeding 97% across both test sets, underscoring their potential for future real-time gait analysis and lower-limb assistive technologies. Full article

(This article belongs to the Special Issue Artificial Intelligence Applications in Healthcare and Precision Medicine, 2nd Edition)

► Show Figures

Figure 1

18 pages, 1173 KB

Open AccessArticle

Machine Learning Methods for Predicting Cancer Complications Using Smartphone Sensor Data: A Prospective Study

by Gabrielė Dargė, Gabrielė Kasputytė, Paulius Savickas, Adomas Bunevičius, Inesa Bunevičienė, Erika Korobeinikova, Domas Vaitiekus, Arturas Inčiūra, Laimonas Jaruševičius, Romas Bunevičius, Ričardas Krikštolaitis, Tomas Krilavičius and Elona Juozaitytė

Appl. Sci. 2026, 16(1), 249; https://doi.org/10.3390/app16010249 - 25 Dec 2025

Viewed by 760

Abstract

Complications are frequent in cancer patients and contribute to adverse outcomes and higher healthcare costs, underscoring the need for earlier identification and prediction. This study evaluated the feasibility of using passively generated smartphone sensor data to explore early-warning signals of complications and symptom [...] Read more.

Complications are frequent in cancer patients and contribute to adverse outcomes and higher healthcare costs, underscoring the need for earlier identification and prediction. This study evaluated the feasibility of using passively generated smartphone sensor data to explore early-warning signals of complications and symptom worsening during cancer treatment. A total of 108 patients were continuously monitored using accelerometer, GPS, and screen on/off data collected through the LAIMA application, while symptoms of depression, fatigue, and nausea were assessed every two weeks and complications were confirmed during clinic visits or emergency presentations. Smartphone data streams were aggregated into variables describing activity and sociability patterns. Machine learning models, including Decision Tree, Extreme Gradient Boosting, K-Nearest Neighbors, and Support Vector Machine, were used for complication prediction, and time-series models such as Autoregressive Integrated Moving Average, Holt–Winters, TBATS, Long Short-Term Memory neural network, and General Regression Neural Network were applied to identify early behavioral changes preceding symptom reports. In this exploratory analysis, the ensemble model demonstrated high sensitivity (89%) for identifying complication events. Smartphone-derived behavioral indicators enabled earlier detection of depression, fatigue, and vomiting by about nine days in a subset of patients. These findings demonstrate the feasibility of passive smartphone sensor data as exploratory early-warning signals, warranting validation in larger cohorts. Full article

(This article belongs to the Special Issue Artificial Intelligence Applications in Healthcare and Precision Medicine, 2nd Edition)

► Show Figures

Figure 1

34 pages, 1875 KB

Open AccessArticle

Retinal Tortuosity Biomarkers as Early Indicators of Disease: Validation of a Comprehensive Analytical Framework

by Mowda Abdalla, Maged Habib, Areti Triantafyllou, Heriberto Cuayáhuitl and Bashir Al-Diri

Appl. Sci. 2025, 15(24), 13136; https://doi.org/10.3390/app152413136 - 14 Dec 2025

Viewed by 560

Abstract

Retinal blood vessel tortuosity is a promising early biomarker for diseases such as diabetic retinopathy. However, the lack of a standardized evaluation method hinders its clinical application. This study presents a framework with 50 features, including 32 developed and refined from our prior [...] Read more.

Retinal blood vessel tortuosity is a promising early biomarker for diseases such as diabetic retinopathy. However, the lack of a standardized evaluation method hinders its clinical application. This study presents a framework with 50 features, including 32 developed and refined from our prior unpublished work. All features were tested for sensitivity and scaling to ensure robust performance. To address the influence of blood vessels’ representation on tortuosity estimation, we tested several resampling approaches and proposed the 1-Equidistant Pixel Sampling method (1EPS), which demonstrated accuracy and approximate scale invariance. The framework was evaluated on a public retinal tortuosity dataset, RET-TORT, consisting of 30 arteries and 30 veins ranked in increased tortuosity. Data augmentation expanded the dataset to 330 arteries and veins for improved reliability. Spearman’s rank correlation coefficient analysis revealed resampling variations in tortuosity estimation, with our method outperforming literature and most features favoring arteries. Using the augmented dataset, Gaussian Process Regression achieved near-perfect performance (

R^{2}

= 1.0 for arteries; 0.999 for veins). Feature selection analysis identified artery- and vein-specific features. This work highlights the importance of accurate vessel preprocessing and feature sensitivity to scaling on tortuosity estimation and introduces a scalable, robust framework of 50 hand-crafted features for clinical tortuosity assessment. Full article

(This article belongs to the Special Issue Artificial Intelligence Applications in Healthcare and Precision Medicine, 2nd Edition)

► Show Figures

Figure 1

13 pages, 341 KB

Open AccessArticle

Effects of Shyness and Adiposity on Heart Rate Reactivity to Psychomotor Challenge in Adolescent Athletes: A Laboratory Study with AI-Supported Data Analysis

by Attila Rausz-Szabó, Veronika Vass, Piroska Béki, Beatrix Faragó and Attila Szabo

Appl. Sci. 2025, 15(24), 13026; https://doi.org/10.3390/app152413026 - 10 Dec 2025

Viewed by 1133

Abstract

Background: Elevated heart rate (HR) reactivity to psychomotor challenge mirrors greater proneness to acute stress, which is a disadvantage in competitive sports. This study investigated whether temperament and adiposity predict HR reactivity during a reaction time (RT) task in adolescent athletes, with a [...] Read more.

Background: Elevated heart rate (HR) reactivity to psychomotor challenge mirrors greater proneness to acute stress, which is a disadvantage in competitive sports. This study investigated whether temperament and adiposity predict HR reactivity during a reaction time (RT) task in adolescent athletes, with a focus on identifying their role in psychophysiological vulnerability. Participants and procedure: The participants were 20 adolescent canoe athletes (15 boys, 5 girls; mean age = 14.3 ± 1.88 years). They were volunteers recruited from a canoe club, with the permission of their coaches and parents. The study was conducted in a controlled laboratory setting, where participants underwent anthropometric tests, completed a questionnaire, had a HR monitor fitted, and rested in an armchair until a relatively stable HR (±5 beats per minute) was recorded. Subsequently, their HR was monitored across three 5 min phases: baseline, RT task, and recovery. Reactivity was calculated as the difference between task and recovery, because pre-task HR was influenced by anticipation. Data analyses were performed using AI-assisted and verified Bootstrapped Spearman correlations, Lasso regression with five-fold cross-validation, and stability analysis with 25 repeated cross-validations. Results: Shyness and body fat percentage were positively related to HR reactivity, whereas other temperament traits and RT performance showed no statistically significant associations. The Lasso regression results revealed shyness and adiposity as significant predictors, with their interaction consistently identified as the strongest effect (selected in 76% of models). The independent measures did not affect HR in the recovery phase. Conclusions: Shy adolescents with higher adiposity demonstrate heightened stress responses, as evidenced by HR reactivity, underscoring the importance of addressing stress vulnerability in young athletes and extending this line of inquiry to a broader spectrum of junior athletes. Full article

(This article belongs to the Special Issue Artificial Intelligence Applications in Healthcare and Precision Medicine, 2nd Edition)

► Show Figures

Figure 1

16 pages, 1353 KB

Open AccessArticle

Comparing Artificial Intelligence (ChatGPT, Gemini, DeepSeek) and Oral Surgeons in Detecting Clinically Relevant Drug–Drug Interactions in Dental Therapy

by Subhi Tayeb, Carlo Barausse, Gerardo Pellegrino, Martina Sansavini, Roberto Pistilli and Pietro Felice

Appl. Sci. 2025, 15(23), 12851; https://doi.org/10.3390/app152312851 - 4 Dec 2025

Cited by 4 | Viewed by 1551

Abstract

Patients undergoing oral surgery are frequently polymedicated and preoperative prescriptions (analgesics, corticosteroids, antibiotics) can generate clinically significant drug–drug interactions (DDIs) associated with bleeding risk, serotonin toxicity, cardiovascular instability and other adverse events. This study prospectively evaluated whether large language models (LLMs) can assist [...] Read more.

Patients undergoing oral surgery are frequently polymedicated and preoperative prescriptions (analgesics, corticosteroids, antibiotics) can generate clinically significant drug–drug interactions (DDIs) associated with bleeding risk, serotonin toxicity, cardiovascular instability and other adverse events. This study prospectively evaluated whether large language models (LLMs) can assist in detecting clinically relevant DDIs at the point of care. Five LLMs (ChatGPT-5, DeepSeek-Chat, DeepSeek-Reasoner, Gemini-Flash, and Gemini-Pro) were compared with a panel of experienced oral surgeons in 500 standardized oral-surgery cases constructed from realistic chronic medication profiles and typical postoperative regimens. For each case, all chronic and procedure-related drugs were provided and the task was to identify DDIs and rate their severity using an ordinal Lexicomp-based scale (A–X), with D/X considered “action required”. Primary outcomes were exact agreement with surgeon consensus and ordinal concordance; secondary outcomes included sensitivity for actionable DDIs, specificity, error pattern and response latency. DeepSeek-Chat reached the highest exact agreement with surgeons (50.6%) and showed perfect specificity (100%) but low sensitivity (18%), missing 82% of actionable D/X alerts. ChatGPT-5 showed the highest sensitivity (98.0%) but lower specificity (56.7%) and generated more false-positive warnings. Median response time was 3.6 s for the fastest model versus 225 s for expert review. These findings indicate that current LLMs can deliver rapid, structured DDI screening in oral surgery but exhibit distinct safety trade-offs between missed critical interactions and alert overcalling. They should therefore be considered as decision-support tools rather than substitutes for clinical judgment and their integration should prioritize validated, supervised workflows. Full article

(This article belongs to the Special Issue Artificial Intelligence Applications in Healthcare and Precision Medicine, 2nd Edition)

► Show Figures

Figure 1

18 pages, 2905 KB

Open AccessArticle

Timestamp Supervision for Surgical Phase Recognition Using Semi-Supervised Deep Learning

by Julia de Enciso García, Alba Centeno López, Ángela González-Cebrián, María Paz Sesmero, Araceli Sanchis, Igor Paredes, Alfonso Lagares and Paula de Toledo

Appl. Sci. 2025, 15(23), 12525; https://doi.org/10.3390/app152312525 - 26 Nov 2025

Viewed by 923

Abstract

Surgical Phase Recognition (SPR) enables real-time, context-aware assistance during surgery, but its use remains limited by the cost and effort of dense video annotation. This study presents a Semi-Supervised Deep Learning framework for SPR in endoscopic pituitary surgery, aiming to reduce annotation requirements [...] Read more.

Surgical Phase Recognition (SPR) enables real-time, context-aware assistance during surgery, but its use remains limited by the cost and effort of dense video annotation. This study presents a Semi-Supervised Deep Learning framework for SPR in endoscopic pituitary surgery, aiming to reduce annotation requirements while maintaining performance. A Timestamp Supervision strategy is employed, where only one or two representative frames per phase are labeled. These labels are then propagated, creating pseudo-labels for unlabeled frames using an Uncertainty-Aware Temporal Diffusion (UATD) approach, based on confidence and temporal consistency. Multiple spatial and temporal architectures are evaluated on the PituPhase–SurgeryAI dataset, the largest publicly available collection of endoscopic pituitary surgeries to date, which includes an outside-the-body phase. Despite using less than 3% of the annotated data, the proposed method achieves an F1-score of 0.60 [0.55–0.65], demonstrating competitive performance against previous Supervised approaches in the same context. Removing the recurrent outside-the-body phase reduces misclassification and improves temporal consistency. These results demonstrate that uncertainty-guided Semi-Supervision is a scalable and clinically viable alternative to fully Supervised Learning for surgical workflow analysis. Full article

(This article belongs to the Special Issue Artificial Intelligence Applications in Healthcare and Precision Medicine, 2nd Edition)

► Show Figures

Figure 1

18 pages, 505 KB

Open AccessArticle

Augmenting, Not Replacing: Clinicians’ Perspectives on AI Adoption in Healthcare

by Francesco Sensi, Francesca Lizzi, Andrea Chincarini, Chiara Binelli, Laura Sartori and Alessandra Retico

Appl. Sci. 2025, 15(23), 12405; https://doi.org/10.3390/app152312405 - 22 Nov 2025

Cited by 4 | Viewed by 1565

Abstract

Artificial intelligence (AI) is widely expected to transform healthcare, yet its adoption in clinical practice remains limited. This paper examines the perspectives of Italian clinicians and medical physicists on the drivers of and barriers to AI use. Using an online survey of healthcare [...] Read more.

Artificial intelligence (AI) is widely expected to transform healthcare, yet its adoption in clinical practice remains limited. This paper examines the perspectives of Italian clinicians and medical physicists on the drivers of and barriers to AI use. Using an online survey of healthcare professionals across different domains, we find that efficiency gains—such as reducing repetitive tasks and accelerating diagnostics—are the strongest incentives for adoption. However, trust in AI systems, explainability, and the limited availability of AI tools are major obstacles. Respondents emphasized that AI should augment, not replace, medical expertise, calling for participatory development processes where clinicians are actively involved in the design and validation of decision support tools. At the organizational level, the adoption of AI tools is facilitated by innovation-oriented leadership and sufficient resources, while conservative management and economic constraints hinder implementation. The awareness of regulatory frameworks, including the EU AI Act, is moderate, and many clinicians express the need for targeted training to support safe integration. Our findings suggest that the successful adoption of AI in healthcare will depend on building trust through transparency, clarifying legal responsibilities, and fostering organizational cultures that support collaboration between humans and AI. The role of AI in medicine is therefore best understood as a complement to clinical judgment, rather than a replacement. Full article

(This article belongs to the Special Issue Artificial Intelligence Applications in Healthcare and Precision Medicine, 2nd Edition)

► Show Figures

Figure 1

15 pages, 1140 KB

Open AccessArticle

Deep Learning Framework for Facial Reconstruction Outcome Prediction: Integrating Image Inpainting and Depth Estimation for Computer-Assisted Surgical Planning

by Fabiano Bini, Guido Manni and Franco Marinozzi

Appl. Sci. 2025, 15(23), 12376; https://doi.org/10.3390/app152312376 - 21 Nov 2025

Viewed by 1032

Abstract

Facial reconstructive surgery requires precise preoperative planning to optimize functional and aesthetic outcomes, but current imaging technologies like CT and MRI do not offer visualization of expected post-surgical appearance, limiting surgical planning capabilities. We developed a deep learning framework integrating facial inpainting and [...] Read more.

Facial reconstructive surgery requires precise preoperative planning to optimize functional and aesthetic outcomes, but current imaging technologies like CT and MRI do not offer visualization of expected post-surgical appearance, limiting surgical planning capabilities. We developed a deep learning framework integrating facial inpainting and monocular depth estimation models to predict surgical outcomes and enable 2D and 3D planning from clinical photographs. Three state-of-the-art inpainting architectures (LaMa, LGNet, MAT) and three monocular depth estimation approaches (ZoeDepth, Depth Anything V2, DepthPro) were evaluated using the FFHQ dataset for inpainting and C3I-SynFace dataset for depth estimation, with comprehensive quantitative metrics assessing reconstruction quality and depth accuracy. For anatomically specific facial features, LGNet demonstrated superior performance across eyebrows (PSNR: 25.11, SSIM: 0.75), eyes (PSNR: 20.08, SSIM: 0.53), nose (PSNR: 25.70, SSIM: 0.88), and mouth (PSNR: 22.39, SSIM: 0.75), with statistically significant differences confirmed by paired t-tests (p < 0.001) and large effect sizes (Cohen’s d = 2.25–6.33). DepthPro significantly outperformed competing depth estimation models with absolute relative difference of 0.1426 (78% improvement over Depth Anything V2: 0.6453 and ZoeDepth: 0.6509) and δ₁ accuracy of 0.8373 (versus 0.6697 and 0.5271 respectively). This novel framework addresses a critical gap in surgical planning by providing comprehensive preoperative visualization of potential outcomes from standard clinical photographs, supporting applications from maxillofacial reconstruction to orbital and nasal procedures. Full article

(This article belongs to the Special Issue Artificial Intelligence Applications in Healthcare and Precision Medicine, 2nd Edition)

► Show Figures

Figure 1

18 pages, 912 KB

Open AccessArticle

Artificial Intelligence in Medicine and Healthcare: A Complexity-Based Framework for Model–Context–Relation Alignment

by Emanuele Di Vita, Giovanni Caivano, Fabio Massimo Sciarra, Simone Lo Bianco, Pietro Messina, Enzo Maria Cumbo, Luigi Caradonna, Salvatore Nigliaccio, Davide Alessio Fontana, Antonio Scardina and Giuseppe Alessandro Scardina

Appl. Sci. 2025, 15(22), 12005; https://doi.org/10.3390/app152212005 - 12 Nov 2025

Cited by 1 | Viewed by 1630

Abstract

Artificial intelligence (AI) is profoundly transforming medicine and healthcare, evolving from analytical tools aimed at automating specific tasks to integrated components of complex socio-technical systems. This work presents a conceptual and theoretical review proposing the Model–Context–Relation (M–C–R) framework to interpret how the effectiveness [...] Read more.

Artificial intelligence (AI) is profoundly transforming medicine and healthcare, evolving from analytical tools aimed at automating specific tasks to integrated components of complex socio-technical systems. This work presents a conceptual and theoretical review proposing the Model–Context–Relation (M–C–R) framework to interpret how the effectiveness of Artificial Intelligence (AI) in medicine and healthcare emerges from the dynamic alignment among algorithmic, contextual, and relational dimensions. No new patient-level data were generated or analyzed. Through a qualitative conceptual framework analysis, the study integrates theoretical, regulatory, and applicative perspectives, drawing on the Revision of the Semiological Paradigm developed by the Palermo School, as well as on major international guidelines (WHO, European AI Act, FDA). The results indicate that AI-supported processes have been reported in the literature to improve clinical accuracy and workflow efficiency when appropriately integrated, yet its value depends on contextual adaptation and human supervision rather than on algorithmic performance alone. When properly integrated, AI functions as a digital semiotic extension of clinical reasoning and may enhance the physician’s interpretative capacity without replacing it. The M–C–R framework enables understanding of how performance, ethical reliability, and organizational sustainability emerge from the alignment between the technical model, the context of use, and relational trust. In this perspective, AI is conceptualized not as a decision-maker but as an adaptive cognitive partner, fostering a reflective, transparent, and person-centered medicine. The proposed approach supports the design of sustainable and ethically responsible AI systems within a Medicine of Complexity, in which human and artificial intelligence co-evolve to strengthen knowledge, accountability, and equity in healthcare systems. Full article

(This article belongs to the Special Issue Artificial Intelligence Applications in Healthcare and Precision Medicine, 2nd Edition)

► Show Figures

Figure 1

15 pages, 2942 KB

Open AccessArticle

Development and Evaluation of a Next-Generation Medication Safety Support System Based on AI and Mixed Reality: A Study from South Korea

by Nathan Lucien Vieira, Su Jin Kim, Sangah Ahn, Ji Sim Yoon, Sook Hyun Park, Jeong Hee Hong, Min-Jeoung Kang, Il Kim, Meong Hi Son, Won Chul Cha and Junsang Yoo

Appl. Sci. 2025, 15(22), 12002; https://doi.org/10.3390/app152212002 - 12 Nov 2025

Viewed by 1647

Abstract

Medication errors pose a significant threat to patient safety. Although Bar-Code Medication Administration (BCMA) has reduced error rates, it is constrained by handheld devices, workflow interruptions, and incomplete safeguards against wrong patients, wrong doses, or drug incompatibility. In this study, we developed and [...] Read more.

Medication errors pose a significant threat to patient safety. Although Bar-Code Medication Administration (BCMA) has reduced error rates, it is constrained by handheld devices, workflow interruptions, and incomplete safeguards against wrong patients, wrong doses, or drug incompatibility. In this study, we developed and evaluated a next-generation BCMA system by integrating artificial intelligence and mixed reality technologies for real-time safety checks: Optical Character Recognition verifies medication–label concordance, facial recognition confirms patient identity, and a rules engine evaluates drug–diluent compatibility. Computer vision models achieved high recognition accuracy for drug vials (100%), medication labels (90%), QR codes (90%), and patient faces (90%), with slightly lower performance for intravenous fluids (80%). A mixed-methods evaluation was conducted in a simulated environment using the System Usability Scale (SUS), Reduced Instructional Materials Motivation Survey (RIMMS), Virtual Reality Sickness Questionnaire (VRSQ), and NASA Task Load Index (NASA-TLX). The results indicated excellent usability (median SUS = 82.5/100), strong user motivation (RIMMS = 3.7/5), minimal cybersickness (VRSQ = 0.4/6), and manageable cognitive workload (NASA-TLX = 31.7/100). Qualitative analysis highlighted the system’s potential to streamline workflow and serve as a digital “second verifier.” These findings suggest strong potential for clinical integration, enhancing medication safety at the point of care. Full article

(This article belongs to the Special Issue Artificial Intelligence Applications in Healthcare and Precision Medicine, 2nd Edition)

► Show Figures

Figure 1

14 pages, 1430 KB

Open AccessArticle

Ensemble-Based Refinement of Landmark Annotations for DNA Ploidy Analysis in Digital Pathology

by Viktor Zoltán Jónás, Dániel Küttel, Béla Molnár and Miklós Kozlovszky

Appl. Sci. 2025, 15(22), 11892; https://doi.org/10.3390/app152211892 - 8 Nov 2025

Viewed by 509

Abstract

Reliable evaluation of image segmentation algorithms in digital pathology depends on high-quality annotation datasets. Landmark-type annotations, essential for cell-counting analyses, are often limited in quality or quantity for segmentation benchmarking, particularly in rare assays where annotation is scarce and costly. In this study, [...] Read more.

Reliable evaluation of image segmentation algorithms in digital pathology depends on high-quality annotation datasets. Landmark-type annotations, essential for cell-counting analyses, are often limited in quality or quantity for segmentation benchmarking, particularly in rare assays where annotation is scarce and costly. In this study, we investigate whether ensemble-inspired refinement of landmark annotations can improve the robustness of segmentation evaluation. Using 15 fluorescently imaged blood samples with more than 20,000 manually placed annotations, we compared three segmentation algorithms—a threshold-based method with clump splitting, a difference-of-Gaussians (DoG) approach, and a convolutional neural network (StarDist)—and used their combined outputs to generate an ensemble-derived ground truth. Confusion matrices and standard metrics (F1 score, precision, and sensitivity) were computed against both manual and ensemble-derived ground truths. Statistical comparisons showed that ensemble-refined annotations reduced noise and decreased mean offsets between annotations and detected objects, yielding more stable evaluation metrics. Our results demonstrate that ensemble-based ground truth generation can guide targeted revision of manual markers, provide a quality measure for annotation reliability, and generate new annotations where no human-generated landmarks exist. This methodology offers a generalizable strategy to strengthen annotation datasets in image cytometry, enabling robust algorithm evaluation in DNA ploidy analysis and potentially in other low-frequency assays. Full article

(This article belongs to the Special Issue Artificial Intelligence Applications in Healthcare and Precision Medicine, 2nd Edition)

► Show Figures

Figure 1

15 pages, 3326 KB

Open AccessArticle

Radiomics and Machine Learning Approaches for the Preoperative Classification of In Situ vs. Invasive Breast Cancer Using Dynamic Contrast-Enhanced Magnetic Resonance Imaging (DCE–MRI)

by Luana Conte, Rocco Rizzo, Alessandra Sallustio, Eleonora Maggiulli, Mariangela Capodieci, Francesco Tramacere, Alessandra Castelluccia, Giuseppe Raso, Ugo De Giorgi, Raffaella Massafra, Maurizio Portaluri, Donato Cascio and Giorgio De Nunzio

Appl. Sci. 2025, 15(14), 7999; https://doi.org/10.3390/app15147999 - 18 Jul 2025

Cited by 2 | Viewed by 1831

Abstract

Accurate preoperative distinction between in situ and invasive Breast Cancer (BC) is critical for clinical decision-making and treatment planning. Radiomics and Machine Learning (ML) have shown promise in enhancing diagnostic performance from breast MRI, yet their application to this specific task remains underexplored. [...] Read more.

Accurate preoperative distinction between in situ and invasive Breast Cancer (BC) is critical for clinical decision-making and treatment planning. Radiomics and Machine Learning (ML) have shown promise in enhancing diagnostic performance from breast MRI, yet their application to this specific task remains underexplored. The aim of this study was to evaluate the performance of several ML classifiers, trained on radiomic features extracted from DCE–MRI and supported by basic clinical information, for the classification of in situ versus invasive BC lesions. In this study, we retrospectively analysed 71 post-contrast DCE–MRI scans (24 in situ, 47 invasive cases). Radiomic features were extracted from manually segmented tumour regions using the PyRadiomics library, and a limited set of basic clinical variables was also included. Several ML classifiers were evaluated in a Leave-One-Out Cross-Validation (LOOCV) scheme. Feature selection was performed using two different strategies: Minimum Redundancy Maximum Relevance (MRMR), mutual information. Axial 3D rotation was used for data augmentation. Support Vector Machine (SVM), K Nearest Neighbors (KNN), Random Forest (RF), and Extreme Gradient Boosting (XGBoost) were the best-performing models, with an Area Under the Curve (AUC) ranging from 0.77 to 0.81. Notably, KNN achieved the best balance between sensitivity and specificity without the need for data augmentation. Our findings confirm that radiomic features extracted from DCE–MRI, combined with well-validated ML models, can effectively support the differentiation of in situ vs. invasive breast cancer. This approach is quite robust even in small datasets and may aid in improving preoperative planning. Further validation on larger cohorts and integration with additional imaging or clinical data are recommended. Full article

(This article belongs to the Special Issue Artificial Intelligence Applications in Healthcare and Precision Medicine, 2nd Edition)

► Show Figures

Figure 1

14 pages, 1520 KB

Open AccessArticle

Machine Learning-Based Detection of Cognitive Impairment from Eye-Tracking in Smooth Pursuit Tasks

by Vida Groznik, Andrea De Gobbis, Dejan Georgiev, Aleš Semeja and Aleksander Sadikov

Appl. Sci. 2025, 15(14), 7785; https://doi.org/10.3390/app15147785 - 11 Jul 2025

Cited by 2 | Viewed by 2009

Abstract

Mild cognitive impairment represents a transitional phase between healthy ageing and dementia, including Alzheimer’s disease. Early detection is essential for timely clinical intervention. This study explores the viability of smooth pursuit eye movements (SPEM) as a non-invasive biomarker for cognitive impairment. A total [...] Read more.

Mild cognitive impairment represents a transitional phase between healthy ageing and dementia, including Alzheimer’s disease. Early detection is essential for timely clinical intervention. This study explores the viability of smooth pursuit eye movements (SPEM) as a non-invasive biomarker for cognitive impairment. A total of 115 participants—62 with cognitive impairment and 53 cognitively healthy controls—underwent comprehensive neuropsychological assessments followed by an eye-tracking task involving smooth pursuit of horizontally and vertically moving stimuli at three different speeds. Quantitative metrics such as tracking accuracy were extracted from the eye movement recordings. These features were used to train machine learning models to distinguish cognitively impaired individuals from controls. The best-performing model achieved an area under the ROC curve (AUC) of approximately 68 %, suggesting that SPEM-based assessment has potential as part of an ensemble of eye-tracking based screening methods for early cognitive decline. Of course, additional paradigms or task designs are required to enhance diagnostic performance. Full article

(This article belongs to the Special Issue Artificial Intelligence Applications in Healthcare and Precision Medicine, 2nd Edition)

► Show Figures

Figure 1

Review

Jump to: Research, Other

19 pages, 1528 KB

Open AccessReview

Deep Learning-Based Prediction of Tumor Mutational Burden from Digital Pathology Slides: A Comprehensive Review

by Dongheng Ma, Hinano Nishikubo, Tomoya Sano and Masakazu Yashiro

Appl. Sci. 2026, 16(3), 1340; https://doi.org/10.3390/app16031340 - 28 Jan 2026

Viewed by 649

Abstract

Tumor mutational burden (TMB) is a key pan-cancer biomarker for immunotherapy selection, but its routine assessment by whole-exome sequencing (WES) or large next-generation sequencing (NGS) panels is costly, time-consuming, and constrained by tissue and DNA quality. In parallel, advances in computational pathology have [...] Read more.

Tumor mutational burden (TMB) is a key pan-cancer biomarker for immunotherapy selection, but its routine assessment by whole-exome sequencing (WES) or large next-generation sequencing (NGS) panels is costly, time-consuming, and constrained by tissue and DNA quality. In parallel, advances in computational pathology have enabled deep learning models to infer molecular biomarkers directly from hematoxylin and eosin (H&E) whole-slide images (WSIs), raising the prospect of a purely digital assay for TMB. In this comprehensive review, we surveyed PubMed and Scopus (2015–2025) to identify original studies that applied deep learning directly to H&E WSIs of human solid tumors for TMB estimation. Across the 17 eligible studies, deep learning models have been applied to predict TMB from H&E WSIs in a variety of tumors, achieving moderate to good discrimination for TMB-high versus TMB-low status. Multimodal architectures tended to outperform conventional CNN-based pipelines. However, heterogeneity in TMB cut-offs, small and imbalanced cohorts, limited external validation, and the black-box nature of these models limit clinical translation. Full article

(This article belongs to the Special Issue Artificial Intelligence Applications in Healthcare and Precision Medicine, 2nd Edition)

► Show Figures

Figure 1

34 pages, 5342 KB

Open AccessReview

Artificial Intelligence in Medical Diagnostics: Foundations, Clinical Applications, and Future Directions

by Dorota Bartusik-Aebisher, Daniel Roshan Justin Raj and David Aebisher

Appl. Sci. 2026, 16(2), 728; https://doi.org/10.3390/app16020728 - 10 Jan 2026

Cited by 2 | Viewed by 3921

Abstract

Artificial intelligence (AI) is rapidly transforming medical diagnostics by allowing for early, accurate, and data-driven clinical decision-making. This review provides an overview of how machine learning (ML), deep learning, and emerging multimodal foundation models have been used in diagnostic procedures across imaging, pathology, [...] Read more.

Artificial intelligence (AI) is rapidly transforming medical diagnostics by allowing for early, accurate, and data-driven clinical decision-making. This review provides an overview of how machine learning (ML), deep learning, and emerging multimodal foundation models have been used in diagnostic procedures across imaging, pathology, molecular analysis, physiological monitoring, and electronic health record (EHR)-integrated decision-support systems. We have discussed the basic computational foundations of supervised, unsupervised, and reinforcement learning and have also shown the importance of data curation, validation metrics, interpretability methods, and feature engineering. The use of AI in many different applications has shown that it can find abnormalities and integrate some features from multi-omics and imaging, which has shown improvements in prognostic modeling. However, concerns about data heterogeneity, model drift, bias, and strict regulatory guidelines still remain and are yet to be addressed in this field. Looking forward, future advancements in federated learning, generative AI, and low-resource diagnostics will pave the way for adaptable and globally accessible AI-assisted diagnostics. Full article

(This article belongs to the Special Issue Artificial Intelligence Applications in Healthcare and Precision Medicine, 2nd Edition)

► Show Figures

Graphical abstract

33 pages, 1546 KB

Open AccessReview

HRV in Stress Monitoring by AI: A Scoping Review

by Giovanna Zimatore, Samuele Russo, Maria Chiara Gallotta, Giordano Passalacqua, Victoria Zaborova, Matteo Campanella, Francesca Fiani, Carlo Baldari, Christian Napoli and Cristian Randieri

Appl. Sci. 2026, 16(1), 23; https://doi.org/10.3390/app16010023 - 19 Dec 2025

Cited by 2 | Viewed by 2390

Abstract

Despite the growing interest in physiological stress monitoring, an objective measure of stress is currently lacking, especially in clinical and rehabilitation contexts. With the emerging integration of artificial intelligence (AI) in data analytics, heart rate variability (HRV) has gained attention as an effective [...] Read more.

Despite the growing interest in physiological stress monitoring, an objective measure of stress is currently lacking, especially in clinical and rehabilitation contexts. With the emerging integration of artificial intelligence (AI) in data analytics, heart rate variability (HRV) has gained attention as an effective biomarker; however, the literature remains fragmented across disciplines, stress types, and methodological approaches. This scoping review aims to investigate how AI techniques are applied to HRV analysis for stress detection and prediction in adult populations. Although this review does not focus on a specific subtype of stress, its primary objective is to explore the current methodological state of the art as reported in the literature, without restrictions on stress typology. Following PRISMA-ScR guidelines, a systematic search was conducted across PubMed, Scopus, and Google Scholar for studies published between 2005 and 2025, using MeSH terms including “HRV”, “Rehabilitation”, “SCI” (for Spinal Cord Injury), “Stress”, “Sympathetic”, “Parasympathetic”, “Non-linear”, “Gamification”, “AI” and “Machine Learning”. Inclusion criteria targeted adult human populations and studies employing HRV features as input for AI and machine learning techniques for psychophysical stress assessment. Of the 566 records identified, 15 studies met the eligibility criteria. The reviewed studies exhibit substantial heterogeneity in terms of settings, populations, sensors, and algorithms with most employing supervised methods (e.g., random forest, support vector machine), alongside several applications of deep learning and explainable AI. Only one study focused specifically on physiological stress, none focused on SCI populations, and rehabilitation-related research was scarce, thereby underscoring important gaps in the current literature. Overall, HR variability analysis, especially when combined with artificial intelligence techniques, represents a promising approach for stress assessment; however, the field is methodologically fragmented and clinically underdeveloped in critical areas, underscoring the need for a multidisciplinary methodological framework. Full article

(This article belongs to the Special Issue Artificial Intelligence Applications in Healthcare and Precision Medicine, 2nd Edition)

► Show Figures

Figure 1

Other

Jump to: Research, Review

25 pages, 1366 KB

Open AccessSystematic Review

Quantifying Readability in Chatbot-Generated Medical Texts Using Classical Linguistic Indices: A Review

by Robert Olszewski, Jakub Brzeziński, Klaudia Watros and Jacek Rysz

Appl. Sci. 2026, 16(3), 1423; https://doi.org/10.3390/app16031423 - 30 Jan 2026

Viewed by 658

Abstract

The rapid development of large language models (LLMs), including ChatGPT, Gemini, and Copilot, has led to their increasing use in health communication and patient education. However, their growing popularity raises important concerns about whether the language they generate aligns with recommended readability standards [...] Read more.

The rapid development of large language models (LLMs), including ChatGPT, Gemini, and Copilot, has led to their increasing use in health communication and patient education. However, their growing popularity raises important concerns about whether the language they generate aligns with recommended readability standards and patient health literacy levels. This review synthesizes evidence on the readability of medical information generated by chatbots using established linguistic readability indices. A comprehensive search of PubMed, Scopus, Web of Science, and Cochrane Library identified 4209 records, from which 140 studies met the eligibility criteria. Across the included publications, 21 chatbots and 14 readability scales were examined, with the Flesch–Kincaid Grade Level and Flesch Reading Ease being the most frequently applied metrics. The results demonstrated substantial variability in readability across chatbot models; however, most texts corresponded to a secondary or early tertiary reading level, exceeding the commonly recommended 8th-grade level for patient-facing materials. ChatGPT-4, Gemini, and Copilot exhibited more consistent readability patterns, whereas ChatGPT-3.5 and Perplexity produced more linguistically complex content. Notably, DeepSeek-V3 and DeepSeek-R1 generated the most accessible responses. The findings suggest that, despite technological advances, AI-generated medical content remains insufficiently readable for general audiences, posing a potential barrier to equitable health communication. These results underscore the need for readability-aware AI design, standardized evaluation frameworks, and future research integrating quantitative readability metrics with patient-level comprehension outcomes. Full article

(This article belongs to the Special Issue Artificial Intelligence Applications in Healthcare and Precision Medicine, 2nd Edition)

► Show Figures

Figure 1

Journal Menu

Journal Browser

Artificial Intelligence Applications in Healthcare and Precision Medicine, 2nd Edition

Share This Special Issue

Special Issue Editors

Special Issue Information

Keywords

Benefits of Publishing in a Special Issue

Published Papers (18 papers)

Research

Review

Other

Further Information

Guidelines

MDPI Initiatives

Follow MDPI