ICT Express最新文献

{"title":"From algorithm to clinic: A critical review of deep learning advances and challenges in dermatological diagnosis and care delivery","authors":"Saima Rahman ,&nbsp;Nisar Uddin ,&nbsp;Hafiz Ishfaq Ahmad ,&nbsp;Moazam Ali ,&nbsp;Sobia Fazal ,&nbsp;Aafaque Ahmad ,&nbsp;Nadia Sultan ,&nbsp;Mashal Fazal Saleem","doi":"10.1016/j.icte.2026.04.010","DOIUrl":"10.1016/j.icte.2026.04.010","url":null,"abstract":"<div><div>This study presents a systematic review of deep learning applications in dermatological diagnosis, analyzing 212 peer-reviewed studies published from 2019 to 2025. A structured search across major scientific databases identified methods, performance trends, and clinical relevance. Advanced architectures such as Vision Transformers, multimodal fusion models, and few-shot learning demonstrate substantial improvements in detecting both common and rare skin conditions. Despite these advances, challenges related to dataset imbalance, clinical integration, and fairness remain significant. The review outlines key research priorities to guide the development of equitable and clinically reliable AI-based dermatological solutions.</div></div>","PeriodicalId":48526,"journal":{"name":"ICT Express","volume":"12 3","pages":"Pages 732-745"},"PeriodicalIF":4.2,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148195680","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Hierarchical adaptive learning and optimization for joint caching and routing in ISL networks","authors":"Jeongmin Seo ,&nbsp;Dongho Ham ,&nbsp;Namseok Ko ,&nbsp;Jeongho Kwak","doi":"10.1016/j.icte.2026.03.018","DOIUrl":"10.1016/j.icte.2026.03.018","url":null,"abstract":"<div><div>With the surge in mobile traffic and the demand for wide coverage, Low-Earth Orbit (LEO) satellite internet has garnered significant attention. However, the limited battery capacity of LEO satellites necessitates efficient energy management. Meanwhile, content caching presents a promising solution to reduce transmission delays and energy consumption of LEO satellites by caching popular files nearby end users. Unlike terrestrial networks, content caching in LEO satellites requires joint optimization with multi-hop routing due to inherent structural differences in ISL (Inter-Satellite Link) networks. In this paper, we formulate an optimization problem aimed at minimizing the average transmission power of LEO satellites while ensuring timely content delivery. To this end, we apply Lyapunov optimization framework to obtain a Hierarchical Caching and Routing (HCR) algorithm where content caching decisions are made over longer periods while routing decisions are made over shorter periods. Finally, via extensive simulations, we demonstrate that the HCR algorithm excels existing algorithms, showing up to an average of 54% and 67% reduction in average queue backlog and average power consumption, respectively.</div></div>","PeriodicalId":48526,"journal":{"name":"ICT Express","volume":"12 3","pages":"Pages 662-668"},"PeriodicalIF":4.2,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148195672","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Continual test-time adaptation in semantic segmentation via confidence-guided adaptive learning","authors":"Junghyeon Seo ,&nbsp;Dahuin Jung ,&nbsp;Jaekoo Lee","doi":"10.1016/j.icte.2026.04.003","DOIUrl":"10.1016/j.icte.2026.04.003","url":null,"abstract":"<div><div>Continual Test-Time Adaptation (CTTA) is essential for adapting models to target data in changing environments while retaining prior knowledge. However, previous methods overlook class imbalance, which limits performance on minor but essential objects. To address this, we propose AdaCoTTA, which applies confidence-guided adaptive learning to improve training stability. These mechanisms mitigate class imbalance, catastrophic forgetting, and error accumulation in CTTA. We evaluate AdaCoTTA on diverse CTTA benchmarks for semantic segmentation. It achieves state-of-the-art performance, improving average mIoU by 0.6% on ACDC and over 4% for minor classes, highlighting its effectiveness in scenarios. Code is available at <span><span>https://github.com/junghyeon0427/AdaCoTTA</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":48526,"journal":{"name":"ICT Express","volume":"12 3","pages":"Pages 714-719"},"PeriodicalIF":4.2,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148195677","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"MA-KANet: Enhancing retinal vessel segmentation through multi-scale feature fusion and attention mechanisms","authors":"Guojie Li ,&nbsp;Fan Zhang ,&nbsp;Muhammad Ateeq ,&nbsp;Anh Nguyen ,&nbsp;Anwar P.P. Abdul Majeed","doi":"10.1016/j.icte.2026.02.006","DOIUrl":"10.1016/j.icte.2026.02.006","url":null,"abstract":"<div><div>We propose Multi-Scale Attention Kolmogorov–Arnold Network (MA-KANet), a novel retinal vessel segmentation framework addressing fine-scale structures and low contrast challenges. Our method integrates Multi-scale Dynamic Fusion (MDF) with 3D convolutional interactions to prevent information loss during downsampling, and Scale-Context Attention Fusion (SCAF) for dynamic feature recalibration in cluttered regions. Kolmogorov–Arnold Network units in decoder stages capture nonlinear dependencies beyond traditional convolutions. MA-KANet achieves state-of-the-art results: 98.80% AUC on DRIVE, 99.05% AUC on CHASEDB1, and 99.17% AUC on FIVES dataset, demonstrating superior performance and generalization across diverse vessel patterns, establishing new benchmarks for retinal vessel segmentation.</div></div>","PeriodicalId":48526,"journal":{"name":"ICT Express","volume":"12 3","pages":"Pages 553-558"},"PeriodicalIF":4.2,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148196076","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A systematic review of handover optimization techniques in 5G self-organizing networks","authors":"Muhammad Umair Munir ,&nbsp;Ismail Ahmedy ,&nbsp;Saaidal Razalli Azzuhri ,&nbsp;Bryan Raj ,&nbsp;Rafidah Md Noor ,&nbsp;Ihsan Ali","doi":"10.1016/j.icte.2026.03.020","DOIUrl":"10.1016/j.icte.2026.03.020","url":null,"abstract":"<div><div>The emergence of 5G technology, along with the rapid expansion of ultra-dense small cell (UDSC) deployments, has significantly increased the complexity of mobility management. Handover (HO) management is a fundamental function of wireless networks that ensures uninterrupted service continuity when a user equipment (UE) moves across different coverage areas or base stations. 5G UDSC environments require efficient handover management for maintaining stable connectivity and high-quality user experience. The effectiveness of handover management depends on the proper adjustment of Handover Control Parameters (HCPs), named as Time-to-Trigger (TTT) and Handover Margin (HOM), these two parameters collectively determine when and how a handover should occur. This study presents a systematic literature review (SLR) that examines handover optimization strategies within 5G Self-Organizing Networks (SON). This study follows the guidelines provided by Kitchenham and the PRISMA framework, with coverage of publications from 2015 to 2025. In this research HO is categorized into three main categories namely, traditional, AI/ML-based, and hybrid approaches. Traditional methods include rule-based, fuzzy logic based, velocity-based, and weighted based techniques that give predictable and efficient solutions, but these approaches are limited by static thresholds and reduced adaptability in dynamic network scenarios. AI/ML-based methods such as supervised learning, reinforcement learning, and deep learning introduce adaptive and predictive tuning of HCPs parameters that result in reductions of handover failures and ping-pong effects and also improve throughput, however, these methods require extensive datasets and computational resources. Hybrid approaches combine the stability of traditional techniques with the adaptability of AI/ML that results in outperforming of single method solutions but still face challenges such as higher complexity and lack of standardization. Most reviewed studies rely on simulation environments such as MATLAB and ns-3, with only a few validated through real-world experimentation or focused specifically on UDSC settings. Overall, this research leads towards the future direction on adaptive, context-aware, and scalable frameworks, emphasizing the need for online learning, hybrid AI integration, privacy-preserving optimization, and large-scale real-world testing to advance intelligent handover management in next-generation networks.</div></div>","PeriodicalId":48526,"journal":{"name":"ICT Express","volume":"12 3","pages":"Pages 669-692"},"PeriodicalIF":4.2,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148195673","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"DropBlock: A structured regularisation technique for mitigating membership inference attacks (MIAs) in machine learning models","authors":"Auwal Shehu Ali ,&nbsp;Zarul Fitri Zaaba ,&nbsp;Manmeet Mahinderjit Singh ,&nbsp;Mohd Ridzuan M Shariff","doi":"10.1016/j.icte.2026.01.006","DOIUrl":"10.1016/j.icte.2026.01.006","url":null,"abstract":"<div><div>This study proposes DropBlock, a structured regularization technique, to mitigate Membership Inference Attacks (MIAs). Unlike traditional dropout, DropBlock masks contiguous regions of feature maps during training, disrupting spatial correlations and preventing data memorization. Evaluated on MNIST, CIFAR-10, and CIFAR-100, DropBlock significantly outperforms baseline defenses, reducing MIA success rates by up to 12.8 % on CIFAR-10 and 12.1 % on CIFAR-100. Furthermore, it suppresses the attack True Positive Rate by over 40 % at low false alarm rates while improving classification accuracy by 2–3 %. These findings confirm that DropBlock provides a superior and practical defense without requiring noise injection.</div><div>2018 The Korean Institute of Communications and Information Sciences. Publishing Services by Elsevier B.V. This is an open access article under the CC BY-NC<img>ND license (<span><span>http://creativecommons.org/licenses/by-nc-nd/4.0/</span><svg><path></path></svg></span>).</div></div>","PeriodicalId":48526,"journal":{"name":"ICT Express","volume":"12 3","pages":"Pages 545-552"},"PeriodicalIF":4.2,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148196075","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Task-adaptive vision experts routing via competency learning guided by predictive uncertainty","authors":"Donghyun Han ,&nbsp;Yuseok Bae ,&nbsp;Jung Uk Kim ,&nbsp;Hyung-Il Kim","doi":"10.1016/j.icte.2026.04.007","DOIUrl":"10.1016/j.icte.2026.04.007","url":null,"abstract":"<div><div>Large-scale pre-trained vision models such as ViT, CLIP, and SAM provide strong foundations for diverse vision tasks, motivating recent Mixture-of-Experts (MoE) approaches that combine multiple experts. However, existing methods often rely on static or implicit routing strategies, limiting adaptability to task semantics and input characteristics. We propose a task-adaptive vision expert routing framework based on competency learning guided by predictive uncertainty. We define expert competency as the relative reduction in predictive uncertainty induced by inter-expert interaction, and formulate expert routing as a learning problem driven by this signal. Our method uses task embeddings derived from textual descriptions to guide expert routing, refines expert features through cross-expert interaction, and aggregates them adaptively into a unified representation. By directly optimizing routing and feature composition using an uncertainty-based competency signal, the model learns how expert collaboration improves task-specific prediction reliability. Extensive experiments on diverse vision tasks demonstrate superior generalization performance and adaptive routing behavior aligned with task semantics.</div></div>","PeriodicalId":48526,"journal":{"name":"ICT Express","volume":"12 3","pages":"Pages 701-706"},"PeriodicalIF":4.2,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148195675","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"RMG-SRGAN:Super-Resolution Generative Adversarial Network based on multi-scale attention aggregation and feature enhancement for radio map generation","authors":"Weizhe Gao,&nbsp;Haini Huang,&nbsp;Jian Wu,&nbsp;Shengbo Hu","doi":"10.1016/j.icte.2026.04.008","DOIUrl":"10.1016/j.icte.2026.04.008","url":null,"abstract":"<div><div>Radio maps are essential for electromagnetic spectrum awareness, supporting communication network optimization and spectrum management. However, existing methods often lack sufficient accuracy in complex propagation environments. To address this, we propose RMG-SRGAN, an improved radio map generation network based on an enhanced super-resolution generative adversarial network. The generator incorporates a Multi-scale Attention Aggregation (MAA) module that strengthens feature representation using multi-scale fusion and dual-path attention in spatial and channel dimensions. The discriminator includes a Feature Enhancement (FE) module to boost discriminative power through multi-stage feature processing. In our evaluation, we prioritize physical fidelity and structural reliability over generic perceptual metrics. Consequently, we employ Root Mean Square Error (RMSE) to quantify the precision of predicted signal strength and the F1-Score to assess the classification accuracy of coverage zones versus blind spots. Extensive experiments on the RadioMapSeer dataset demonstrate that RMG-SRGAN achieves state-of-the-art performance, securing the lowest RMSE and highest F1-Score compared to existing baselines.</div></div>","PeriodicalId":48526,"journal":{"name":"ICT Express","volume":"12 3","pages":"Pages 720-725"},"PeriodicalIF":4.2,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148195678","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Physical-layer security enhancement in cognitive radio networks using whale optimization-based sidelobe suppression","authors":"Atif Elahi ,&nbsp;Noor Gul ,&nbsp;Adnan Elahi ,&nbsp;Su Min Kim ,&nbsp;Junsu Kim","doi":"10.1016/j.icte.2026.04.009","DOIUrl":"10.1016/j.icte.2026.04.009","url":null,"abstract":"<div><div>OFDM-based cognitive radio networks suffer from high sidelobe emissions, leading to adjacent interference and potential security vulnerabilities. This paper introduces a Whale Optimization Algorithm (WOA)-based cancellation carrier (CC) approach for sidelobe suppression. Unlike conventional fixed-amplitude methods, the proposed framework adaptively optimizes CC amplitudes under power constraints to minimize total sidelobe energy (TSE). Simulation results show significant improvements: Peak Sidelobe Level (PSL) reduces from −41.48 dB to −57.1 dB, TSE drops from −57.60 dB to −80 dB, and Adjacent Channel Leakage Ratio (ACLR) approaches zero, while maintaining stable Bit Error Rate (BER) and improved Peak-to-Average Power Ratio (PAPR). These improvements enhance spectral efficiency and improve physical-layer robustness by reducing spectral leakage, which may limit susceptibility to eavesdropping, jamming, and spoofing.</div></div>","PeriodicalId":48526,"journal":{"name":"ICT Express","volume":"12 3","pages":"Pages 726-731"},"PeriodicalIF":4.2,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148195679","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"GluPatchNet: A Channel-Mixed and decomposition-based blood glucose prediction scheme for mitigating CGM time delay","authors":"Su-Jin Kim ,&nbsp;Sungho Kim ,&nbsp;Jun Sung Moon ,&nbsp;Sung-Yoon Jung","doi":"10.1016/j.icte.2026.04.011","DOIUrl":"10.1016/j.icte.2026.04.011","url":null,"abstract":"<div><div>The Artificial Pancreas System (APS) is a device that continuously monitors the blood glucose levels of diabetic patients in real-time and automatically adjusts insulin delivery based on CGM data. However, glucose data obtained from Continuous Glucose Monitoring (CGM) sensors typically exhibit a time delay of approximately 10 min relative to actual blood glucose levels measured by Blood Glucose Monitoring (BGM). This time delay can hinder the timely administration of insulin. To address this issue, blood glucose levels can be predicted in advance to compensate for the delay, with the predicted values complementing real-time data to improve the accuracy and efficiency of the APS. In this paper, we propose GluPatchNet, an enhanced blood glucose prediction model based on PatchTST, utilizing various combinations of input features such as blood glucose (BG), carbohydrate intake (CHO), and insulin administration. GluPatchNet combines Channel-Mixed operations with Residual-Trend Decomposition techniques to effectively model both long-term trends and short-term variations. We evaluated the blood glucose prediction performance using data generated from the UVA-Padova virtual simulator through Simglucose. Simulation results show that GluPatchNet outperforms the original PatchTST and other deep learning models, achieving the highest prediction accuracy in both single-input and multi-input scenarios.</div></div>","PeriodicalId":48526,"journal":{"name":"ICT Express","volume":"12 3","pages":"Pages 746-751"},"PeriodicalIF":4.2,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148195650","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}