Saved Queries

With the rapid growth of fashion e-commerce, fashion attribute recognition has emerged as a critical research area in computer vision. Existing methods face two primary problems: (1) building multi-task models, leading to complex network architectures; (2) the overlooking of semantic relationships and spatial positional dependencies between fashion attributes. To address these issues, this paper proposes SLAR-Net, a novel hierarchical multi-label classification network that effectively fuses spatial and semantic information for improved recognition performance. Specifically, SLAR-Net adopts a progressive, hierarchical architecture. Firstly, we introduce a lightweight backbone network enhanced with a custom-designed attention mechanism to extract low-level image features. Secondly, we innovatively construct an adjacency matrix to represent the relative spatial orientations of attributes, which is then employed by a graph convolutional network to model mid-level spatial positional features. Thirdly, we design a graph embedding matrix that captures attribute dependency relationships, leveraging a neural network to learn high-level semantic representations. Finally, we propose a custom multi-head attention mechanism to fuse spatial and semantic features, facilitating enhanced feature interaction and improving recognition performance. Experimental results on fashion attribute and benchmark datasets demonstrate that SLAR-Net outperforms state-of-the-art methods in recognition accuracy, validating the effectiveness of the proposed hierarchical architecture and fusion strategy. Full article

(This article belongs to the Special Issue Advanced Technologies in Image Processing, Analysis, and Machine Vision)

►▼ Show Figures

Figure 1

24 pages, 3066 KB

Open AccessArticle

Enhancing Network Traffic Monitoring Through eXplainable Artificial Intelligence Methodologies

by Cătălin-Eugen Bucur, Georgiana Crihan, Anamaria Rădoi, Elena-Grațiela Robe-Voinea and Iustin-Nicolae Moroșan

Telecom 2026, 7(2), 34; https://doi.org/10.3390/telecom7020034 (registering DOI) - 23 Mar 2026

Abstract

In the contemporary digital landscape, AI (Artificial Intelligence) emerged as a pivotal tool in enhancing the defense technologies developed across the entire network infrastructure. As reliance on AI-based decision-making grew, so did the imperative need for interpretability, transparency, and trustworthiness, leading to the development and integration of XAI (eXplainable Artificial Intelligence). This research paper provides a comprehensive overview of the current state of the art in XAI approaches that can be effectively implemented for network traffic monitoring, especially in critical digital infrastructures. The main contribution of this research article consists of the comparative analysis of the XAI SHAP (Shapley Additive Explanation) method applied to different datasets obtained from real-time network traffic monitoring, utilizing several representative parameters, which demonstrates the performance, vulnerabilities, and limitations of the proposed method, and also the security implications of the system resources from a cybersecurity perspective. Experimental results show that Ethernet networks offer higher predictability and clearer decision boundaries. Consequently, they are a safer solution for deployment in sensitive network architectures. In contrast, BYOD (Bring Your Own Device) Wi-Fi environments exhibit greater randomness. Full article

►▼ Show Figures

Figure 1

22 pages, 2820 KB

Open AccessArticle

Designing Visual Arts Education for Sustainability: An Arts-Based Approach to Fostering Ecological Awareness in Pre-Service Teachers

by Zlata Tomljenović

Sustainability 2026, 18(6), 3131; https://doi.org/10.3390/su18063131 (registering DOI) - 23 Mar 2026

Abstract

Visual arts education (VAE) offers a promising pedagogical space for addressing sustainability challenges by engaging the cognitive, emotional, and ethical dimensions of learning. This study examines how engagement with contemporary visual arts and art-based pedagogical practices can foster ecological thinking, ecological literacy, and sustainability awareness among pre-service teachers. The research was conducted over one academic year (2022/2023) within two visual arts courses attended by a total of 69 second- and third-year students enrolled in a teacher education programme. Using a qualitative, interpretative research design, the study investigated how selected contemporary artworks addressing ecological themes were pedagogically contextualised and discussed, and how students engaged with these artworks through dialogue, reflection, and their own art-making processes. Data were collected from students’ written reflections, group discussions, and visual works, and analysed using an interpretative framework informed by visual hermeneutics and sustainability education discourse. The findings indicate that engagement with contemporary visual art can foster the development of ecological literacy by enabling students to integrate experiential, affective, reflective, and relational dimensions of sustainability into their understanding of environmental issues. In line with the objectives of SDG 4 (Quality Education), SDG 12 (Responsible Consumption and Production), and SDG 13 (Climate Action), the study contributes to existing literature by demonstrating the pedagogical potential of visual arts education within teacher education and Education for Sustainable Development. Full article

►▼ Show Figures

Figure 1

30 pages, 4202 KB

Open AccessArticle

Study on Post-Use Evaluation and Optimization Strategies for the Cultural Tourism Landscape of Xidajie Street in Baoding from the Perspective of Immersive Experience

by Ke Ni, Ji Feng, Chenyu Wang, Yanwei Zhou and Heng Wang

Buildings 2026, 16(6), 1259; https://doi.org/10.3390/buildings16061259 - 23 Mar 2026

Abstract

In the context of immersive technologies deeply integrated into the cultural tourism industry, immersive cultural tourism has become an important means of heritage revitalization. Immersive experience is both a crucial consumer element and a key indicator for evaluating the attractiveness of cultural tourism landscapes. This study evaluates the post-use experience of the cultural tourism landscape of Xidajie Street in Baoding from the perspective of tourist immersion. Through a literature review, investigation of typical immersive districts, and expert interviews, we extract immersive cultural tourism landscape evaluation criteria based on a depth model of immersion, focusing on three dimensions: narrative, enclosure, and interaction. Subjective perception data from tourists is then collected through a survey, and IPA (Importance–Performance Analysis) is employed to identify the strengths and weaknesses of Xidajie’s cultural tourism landscape. The results show that Xidajie excels in spatial environment shaping and historical preservation, but has room for improvement in cultural narrative extension, contextual immersion, and interactive experiences. Therefore, strategies are proposed to enhance the cultural IP, establish a complete narrative structure, create authentic enveloping environments, and enrich interactive games to build a high-quality online and offline immersive cultural tourism landscape. This aims to promote the renewal of Xidajie and the dynamic transmission of Baoding’s local culture. Full article

(This article belongs to the Special Issue Advanced Study on Urban Environment by Big Data Analytics)

►▼ Show Figures

Figure 1

31 pages, 1926 KB

Open AccessArticle

FairAgent: A Collaborative Multi-Agent System for Fair Competition Review

by Yuanqing Mao, Jinfei Ye, Cheng Yang, Chuncong Wang, Qiyu Chen, Yang Xu, Min Zhu, Hanrui Chen, Jiong Lin, Beining Wu and Feiwei Qin

Electronics 2026, 15(6), 1329; https://doi.org/10.3390/electronics15061329 - 23 Mar 2026

Abstract

The rapid progress of large language models (LLMs) has fostered the development of domain-specific variants in law, medicine, and finance. However, existing legal LLMs still struggle to generate contextually grounded and regulation-compliant responses in complex scenarios of fair competition review. To address this, we present FairAgent, a collaborative multi-agent framework that unifies data refinement and reinforcement learning for legal reasoning. FairAgent integrates two core modules: (1) EchoCourt, a closed-loop data generation and refinement pipeline that constructs high-quality question–answer pairs through generation, critique, and optimization guided by a hierarchical Fairness Knowledge Forest; and (2) a two-stage outcome-based reinforcement learning mechanism that progressively teaches the model to invoke and integrate external retrieval in reasoning. We further enhance learning stability through a RAG-based rollout and retrieval-mask loss. Extensive evaluations demonstrate that FairAgent significantly improves reasoning accuracy, interpretability, and compliance in fair competition review compared with state-of-the-art baselines, establishing a scalable framework for retrieval-augmented legal intelligence. Full article

(This article belongs to the Special Issue AI-Driven Natural Language Processing Applications)

►▼ Show Figures

Figure 1

16 pages, 1676 KB

Open AccessArticle

Multimodal Bone Fragility Profiling in People Living with HIV: Trabecular Bone Score, Calcaneal Quantitative Ultrasound, and Sarcopenia Screening

by David Vladut Razvan, Jenel Marian Patrascu, Ovidiu Rosca, Iulia Georgiana Bogdan, Livia Stanga, Adrian Vlad and Camelia Vidita Gurban

Medicina 2026, 62(3), 603; https://doi.org/10.3390/medicina62030603 (registering DOI) - 23 Mar 2026

Abstract

Background and Objectives: Bone fragility in people living with HIV (PLWH) reflects both reduced bone mineral density (BMD) and impaired microarchitecture, while functional decline may further amplify fracture vulnerability. This study evaluated whether adding a pragmatic sarcopenia screen improves bone fragility characterization beyond DXA-BMD, trabecular bone score (TBS), calcaneal quantitative ultrasound (QUS), and biomarkers, and explored the relationship between tenofovir disoproxil fumarate (TDF) exposure and microarchitectural impairment. Materials and Methods: In this single-center cross-sectional study at Victor Babeș University of Medicine and Pharmacy Timișoara, 98 adults on stable ART underwent DXA (T-scores), lumbar TBS (reported as TBS × 100), calcaneal QUS (SOS/BUA), and bone turnover markers (CTX, P1NP, 25(OH)D). Sarcopenia screening used handgrip strength and 4 m gait speed. Associations were tested using group comparisons, correlations, and multivariable modeling for degraded TBS (TBS × 100 < 124.0). Results: Sarcopenia screen-positive participants (n = 28) had lower TBS (123.8 vs. 127.7, p = 0.02), lower lumbar T-score (−1.7 vs. −1.2, p = 0.014), lower SOS (1523.3 vs. 1548.8 m/s, p = 0.002), and higher CTX (0.6 vs. 0.4 ng/mL, p < 0.001), with less frequent viral suppression (60.7% vs. 85.7%, p = 0.006). With >5 years TDF exposure (n = 28), degraded TBS prevalence was 82.1% vs. 40.0% in never-exposed (p = 0.001), alongside lower TBS (123.1 vs. 129.8, p < 0.001) and higher CTX (0.6 vs. 0.4 ng/mL, p < 0.001). Viral suppression independently reduced odds of degraded TBS (aOR 0.3, 95% CI 0.1–0.9; p = 0.034). Conclusions: In PLWH, prolonged TDF exposure and functional impairment co-occur with worse densitometric and microarchitectural profiles; viral suppression shows an independent protective association with microarchitecture. Full article

(This article belongs to the Section Orthopedics)

►▼ Show Figures

Figure 1

19 pages, 893 KB

Open AccessArticle

A Bayesian Framework with Dirichlet Priors and Spatial Smoothing for Protein Rotamer Prediction

by Kamal Al Nasr, Ahmad Jad Allah, Mohammad Alamri and Mohammad Al Sallal

Int. J. Mol. Sci. 2026, 27(6), 2869; https://doi.org/10.3390/ijms27062869 - 22 Mar 2026

Abstract

Accurate prediction of protein sidechain conformations is a fundamental challenge in structural biology, with diverse applications ranging from protein structure determination to computational drug design. The performance of backbone-dependent rotamer libraries is often limited by discrete binning artifacts and difficulties handling sparse conformational regions. In this work, we present a Bayesian framework for rotamer prediction that addresses these limitations through Dirichlet priors and spatial smoothing. Our approach models rotamer probabilities as continuous functions of backbone dihedral angles, using circular Gaussian convolution, to make the most of statistical strength from neighboring conformations while respecting the periodic nature of angular data. We constructed rotamer libraries through structural clustering of sidechain conformations rather than chi angle binning, ensuring that each rotamer represents a distinct three-dimensional geometry. We evaluated and compared our framework against the state-of-the-art libraries on two independent test sets. Our Dirichlet model achieved chi angle prediction accuracy of 59–60%. Notably, our method produced consistently lower angular errors, an approximate 13% reduction in mean deviation, suggesting that the continuous probability distributions better capture subtle conformational preferences. Further, we explored the incorporation of non-sequential context by including the identity of nearby non-neighboring residues as an example of extensibility of our framework. Full article

(This article belongs to the Section Molecular Biophysics)

9 pages, 664 KB

Open AccessReview

The Inflammatory, Apoptotic, and Cardiovascular Role of Soluble and Tissue Gp120 in PLWH on Antiretroviral Therapy: Is Anti-gp120 Therapy Needed?

by Alessia Mirabile, Dalida Bivona, Giuseppe Nicolò Conti, Andrea Marino, Benedetto Maurizio Celesia, Grazia Scuderi, Paolo Fagone, Serena Matera, Serena Spampinato and Giuseppe Nunnari

Acta Microbiol. Hell. 2026, 71(1), 8; https://doi.org/10.3390/amh71010008 (registering DOI) - 22 Mar 2026

Abstract

People living with HIV (PLWH) receiving effective antiretroviral therapy (ART) continue to exhibit chronic immune activation and systemic inflammation despite virological suppression. The viral envelope glycoprotein gp120, which binds the CD4 receptor and mediates viral entry, has been implicated in pro-inflammatory and pro-apoptotic effects in neuronal and endothelial cells. Although gp120 is expressed on the viral surface, its oligomeric structure and its ability to form immune complexes with circulating antibodies may reduce the sensitivity of standard detection assays in serum. Soluble gp120 has been associated with increased levels of pro-inflammatory cytokines, including interleukin-6 (IL-6), tumor necrosis factor-α (TNF-α), and interleukin-1β (IL-1β), as well as chemokines. These mediators may contribute to blood–brain barrier dysfunction, endothelial injury, vascular smooth muscle alterations, and subsequent neurodegenerative and cardiovascular complications. Importantly, gp120 shedding may persist due to viral reservoirs and intermittent reactivation, even during ART. Fostemsavir inhibits the interaction between gp120 and CD4, preventing viral entry and potentially limiting gp120-mediated pathogenic effects. Beyond antiviral activity, this mechanism suggests a potential role in attenuating gp120-mediated inflammation. This review discusses the biological effects of gp120 and the rationale for targeting it therapeutically in PLWH. Full article

►▼ Show Figures

Figure 1

19 pages, 13646 KB

Open AccessArticle

CA-GFNet: A Cross-Modal Adaptive Gated Fusion Network for Facial Emotion Recognition

by Sitara Afzal and Jong-Ha Lee

Mathematics 2026, 14(6), 1068; https://doi.org/10.3390/math14061068 - 21 Mar 2026

Abstract

Facial emotion recognition (FER) plays an important role in healthcare, human–computer interaction, and intelligent security systems. However, despite recent advances, many state-of-the-art FER methods depend on computationally intensive CNN or transformer backbones and large-scale annotated datasets while suffering noticeable performance degradation under cross-dataset evaluation because of domain shift. These limitations hinder practical usage in resource-constrained and real-world environments. To address this issue, we propose Cross-Adaptive Gated Fusion Network (CA-GFNet), a lightweight dual-stream FER framework that explicitly combines shallow structural features with deep semantic representations. The proposed architecture integrates domain-robust gradient-based descriptors with compact deep features extracted from a VGG-based backbone. After face detection and normalization, the structural stream captures fine-grained local appearance cues, whereas the semantic stream encodes high-level facial configurations. The two feature streams are projected into a shared latent space and adaptively fused using a gated fusion mechanism that learns sample-specific weights, allowing the model to prioritize the more reliable feature source under dataset shift. Extensive experiments on KDEF along with zero-shot cross-dataset evaluation on CK+ using a strict train-on-KDEF/test-on-CK+ protocol with subject-independent splits demonstrate the effectiveness of the proposed method. CA-GFNet achieves 99.30% accuracy on KDEF and 98.98% on CK+ while requiring significantly fewer parameters than conventional deep FER models. These results confirm that adaptive gated fusion of shallow and deep features can deliver both high recognition accuracy and strong cross-dataset robustness. Full article

(This article belongs to the Special Issue Advanced Algorithms in Multimodal Affective Computing)

19 pages, 37748 KB

Open AccessArticle

Factually Consistent Prompting with LLMs for Cross-Lingual Dialogue Summarization

by Zhongtian Bao, Wenjian Ding, Yao Zhang, Jun Wang, Zhe Sun, Andrzej Cichocki and Zhenglu Yang

Computers 2026, 15(3), 197; https://doi.org/10.3390/computers15030197 - 21 Mar 2026

Abstract

Recent breakthroughs in large language models have made it feasible to effectively summarize cross-lingual dialogue information, proving essential for the global communication context. However, existing methodologies encounter difficulties in maintaining factual consistency across multiple dialogue exchanges and lack clear explanations of the summarization process. This paper presents a novel factually consistent prompting technology with large language models to address these challenges in cross-lingual dialogue summarization. First, we propose a factual replacement mechanism to enhance information analysis by incorporating noise information into summarization candidates. We adopt a self-guidance framework to enforce factual consistency, enhancing information flow tracking in cross-lingual hybrid dialogue scenarios with the assistance of GPT-based models. Furthermore, we introduce a view-aware chain-of-thought-driven architecture to improve the interpretability and transparency of the cross-lingual dialogue summarization process. Comprehensive experimental evaluations on cross-lingual summarization tasks, spanning English, French, Spanish, Russian, Chinese, and Arabic, and hybrid cross-lingual tasks substantiate that the proposed model achieves superior performance relative to state-of-the-art baselines. Full article

(This article belongs to the Topic Artificial Intelligence Models, Tools and Applications: 2nd Edition)

35 pages, 710 KB

Open AccessReview

AI Agent Communications in the Future Internet—Paving a Path Toward the Agentic Web

by Qiang Duan and Zhihui Lu

Future Internet 2026, 18(3), 171; https://doi.org/10.3390/fi18030171 - 21 Mar 2026

Abstract

The rapid evolution of artificial intelligence technologies toward the agentic AI paradigm enables the emergence of the Agentic Web in the future Internet. Agent communication plays a critical role in constructing the Agentic Web but faces unique challenges posed by the edge–network–cloud continuum in the future Internet. This paper provides a comprehensive overview of state-of-the-art agent communication protocols and technologies, evaluating their readiness to support the construction of the Agentic Web. We first survey representative communication protocols and analyze the key technologies they employ, assessing their effectiveness in addressing the challenges for agent communications in the future Internet. We then identify critical gaps between existing approaches and the requirements of the Agentic Web, and propose a unified architectural framework grounded in virtualization and service-oriented principles to address these gaps. Such a framework may greatly facilitate the development of a pluralistic ecosystem in which various agent communication technologies and protocols can be freely developed and fully utilized. We also discuss open topics and possible directions for future research toward a fully realized Agentic Web. Full article

(This article belongs to the Special Issue 2026 and 2027 Feature Papers from Future Internet’s Editorial Board Members)

►▼ Show Figures

Figure 1

13 pages, 2876 KB

Open AccessArticle

A Frequency-Aware Self-Supervised Framework for MEMS-OCT Denoising

by Gaolin Zhang, Zonghao Li, Hui Zhao, Zhe Peng and Huikai Xie

Biosensors 2026, 16(3), 177; https://doi.org/10.3390/bios16030177 - 21 Mar 2026

Abstract

Optical coherence tomography (OCT) is a key biological sensing and imaging tool widely used in biomedical detection, and its images are often degraded by multiplicative speckle noises—especially when micro-electro-mechanical system (MEMS) mirrors are employed in endoscopic OCT imaging, which reduces visual quality and affects the accuracy of subsequent analysis. Traditional denoising algorithms and supervised deep learning approaches have shown some effectiveness, but they are limited by their reliance on paired noisy–clean data and their insufficient modeling of global structural dependencies. To address these issues, this paper proposes a frequency-domain enhanced UNet based on the Neighbor2Neighbor (N2N) framework (FEN2N). The proposed FEN2N integrates wavelet-guided spectral pooling modules (WSPMs) and frequency-domain enhanced receptive field blocks (FE-RFBs). In this work, OCT images are obtained in a self-constructed MEMS-OCT system. Then the FEN2N is applied to the OCT image dataset. Results show that FEN2N achieves a more than 2.3 dB PSNR improvement over the N2N baseline, while the incorporation of FE-RFB contributes to a 0.02 improvement in SSIM. In addition, FEN2N outperforms several state-of-the-art methods, effectively suppressing speckle noise while preserving fine structural details that are important for clinical diagnosis. Full article

(This article belongs to the Section Optical and Photonic Biosensors)

14 pages, 456 KB

Open AccessArticle

Predictors of Late Adverse Outcomes After Carotid Endarterectomy

by Danka Vukasinovic, Milos Maksimovic, Slobodan Tanaskovic, Jelena Marinkovic, Andja Cirkovic, Branko Jakovljevic, Jelena Ilic Zivojinovic, Djordje Radak and Hristina Vlajinac

Medicina 2026, 62(3), 593; https://doi.org/10.3390/medicina62030593 (registering DOI) - 21 Mar 2026

Abstract

Background and Objectives: Although carotid endarterectomy (CEA) is the gold standard in the treatment of carotid disease, a higher frequency of adverse outcomes can reduce its benefit. The aim of the present study is to identify factors related to myocardial infarction, stroke, death and restenosis as the late adverse outcomes of CEA. Materials and Methods: The retrospective cohort study included 1597 CEAs that were performed in 1533 consecutive patients at the Vascular Surgery Clinic in Belgrade from 2012 to 2017. Late adverse outcomes within 4 years after CEA were available for the majority of them. Data for myocardial infarction and stroke were available for 1223 CEAs, data for death for 1305 CEAs, and data for restenosis for 1162 CEAs. The association between possible risk factors and late adverse outcomes of CEA was analyzed using univariate and multivariate Cox and logistic regression analyses. Results: During follow-up, myocardial infarction occurred after 55, stroke after 68, death after 103 and restenosis after 121 CEAs. Two factors were the most frequent predictors of late adverse outcomes, i.e., the patient’s age and diabetes mellitus (DM). Age predicted all late adverse outcomes except restenosis, and DM predicted all of them. A predictor of myocardial infarction, besides age (HR 1.08, 95% CI 1.05–1.11) and DM (HR 1.60, 95% CI 1.11–2.29), was peripheral arterial disease (HR 1.81, 95% CI 1.17–2.78) in personal history. Predictors were only age (HR 1.04, 95% CI 1.01–1.08) and DM (HR 1.68, 95% CI 1.03–2.72) for stroke, as well as for death (HR 1.17, 95% CI 1.12–1.21 and HR 1.94, 95% CI 1.17–3.21, respectively). For restenosis, in addition to DM (HR 1.78, 95% CI 2.62), predictors were hyperlipidemia (HR 3.52, 95% CI 1.27–9.76) and urgent surgery (HR 3.51, 95% CI 1.06–11.65). Conclusions: CEA should be performed with special caution in the elderly and diabetic patients. Modification of other risk factors and precise medical therapy are necessary to reduce possible adverse outcomes. Full article

(This article belongs to the Section Surgery)

►▼ Show Figures

Figure 1

31 pages, 3749 KB

Open AccessArticle

Nomadic Gardens as a Design Paradigm: Linking Everyday Practices, Cultural Memory and Adaptive Urbanism

by Sonia Vuscan, Jianglong Yu and Radu Muntean

Sustainability 2026, 18(6), 3107; https://doi.org/10.3390/su18063107 (registering DOI) - 21 Mar 2026

Abstract

Rapid, state-led urbanization in China often generates socio-spatial vulnerabilities, leaving interstitial “waiting lands” in a state of regulatory and ecological limbo. This paper investigates “nomadic gardens”—spontaneous, resident-led cultivation in Jinan—as a bottom-up strategy for adaptive capacity. Using a mixed-methods approach involving site typologies and community surveys (n = 100), we identify eight distinct garden forms that function as socio-ecological buffers, mitigating the risks of social isolation and psychological distress among elderly residents. Findings reveal a significant resilience gap caused by rigid land-use policies that prioritize ornamental aesthetics over functional productivity. We propose an Adaptive Urbanism framework that utilizes modular design and transitional governance to transform these precarious spaces into managed resilience assets. By shifting the planning focus from enforcement to risk-responsive design, this research provides a scalable model for sustainable urban risk management in rapidly transforming global cities. Full article

(This article belongs to the Special Issue Sustainable Urban Risk Management and Resilience Strategy)

►▼ Show Figures

Figure 1

29 pages, 1953 KB

Open AccessFeature PaperArticle

JDC-DA: An Unsupervised Target Domain Algorithm for Alzheimer’s Disease Diagnosis with Structural MRI Using Joint Domain and Category Dual Adaptation

by Yuan Sui, Yujie Zhang, Ying Wei and Gang Yang

Mathematics 2026, 14(6), 1067; https://doi.org/10.3390/math14061067 - 21 Mar 2026

Abstract

Domain shift in multi-source MRI imaging data significantly degrades the performance of Alzheimer’s disease diagnostic models. This study aims to develop an effective unsupervised domain adaptation method to enhance diagnostic accuracy across different clinical datasets. We propose a Joint Domain and Category Dual Adaptation framework (JDC-DA) that integrates metric learning and adversarial learning. The method employs multi-scale feature aggregation to capture diverse lesion characteristics, generates dynamic prototype features through category clustering, and implements a novel metric learning approach that simultaneously aligns both domain-level and category-level feature distributions. Additionally, we introduce a classification certainty maximization strategy that establishes a dual adversarial mechanism between domain discriminator and classification discrepancy discriminator. The framework was evaluated on four public datasets (ADNI-1, ADNI-2, ADNI-3, AIBL) containing 1230 baseline sMRI scans for four classification tasks: AD vs. NC, MCI vs. NC, AD vs. MCI, and AD vs. MCI vs. NC. The proposed JDC-DA method achieved superior performance with accuracies of 92.16%, 83.56%, 81.96%, and 79.12% for the four classification tasks respectively, significantly outperforming existing state-of-the-art domain adaptation methods across all evaluation metrics. The JDC-DA framework effectively addresses domain shift challenges in Alzheimer’s disease diagnosis through its integrated approach to feature alignment and adversarial learning. The method demonstrates strong potential for clinical application in automated diagnosis systems, particularly for handling multi-center neuroimaging data with distribution discrepancies. Full article

(This article belongs to the Special Issue Recent Advances in Machine Learning Methods for Medical Imaging Analysis)

►▼ Show Figures

Figure 1

Show export options Show export options

Select all

Export citation of selected articles as:

Error

Oops... you haven't selected anything for export.

Displaying article 1-50 on page 1 of 871.

Go to page 1 2 3 4 5

Search Results (43,505)

Further Information

Guidelines

MDPI Initiatives

Follow MDPI