alexandria engineering journal最新文献_第10页

Quantum-inspired deep learning optimisation for real-time student engagement analysis in virtual classrooms 量子启发的深度学习优化，用于虚拟教室中的实时学生参与分析

IF 6.8 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

alexandria engineering journal

Pub Date : 2025-12-19 DOI: 10.1016/j.aej.2025.12.019

Hanan T. Halawani , Shuhrah Alghamdi , Fahad Ali Aloufi , Maryam Alsolami , Abdulellah Alsulaimani , Hassan M. Aljohani

Quantum computing (QC) employs quantum-mechanical principles such as superposition and entanglement to solve specific problems far more efficiently than classical computers. E-learning is the only option for students and teachers during the pandemic. Conversely, it is challenging for an instructor to observe every student’s engagement while educating online. Students are distracted during such activities. The teachers want to identify their students' states, whether they are concentrated or concerned with teaching. Hence, from a teacher’s viewpoint, it is significant to assess students' levels of engagement to understand their actual reactions and take the necessary steps to involve students and help them achieve goals, machine learning (ML) and deep learning (DL) is deployed for predictive analytics of a student’s performance and engagement depending upon interactions, contribution in class, etc. In this paper, the Deep Learning-Driven Quantum Inspired Moth Flame Optimizer for Real-Time Student Engagement Analysis (DLQIMFO-RSEA) method is proposed. The DLQIMFO-RSEA method aims to categorise student engagement in online classes. To accomplish this, the DLQIMFO-RSEA method uses the YOLOv5 object detection model with backbones SPPF, CBS, and CSPI-X. Next, the image pre-processing stage employs the Wiener filter (WF) to remove the noise. For feature extraction, the InceptionResNetV2 technique is used. Furthermore, a stacked autoencoder (SAE) is applied for detection. At last, the parameter tuning process is performed by the quantum-inspired moth flame optimiser (QIMFO) model to improve the classification performance of the SAE model. The comparison analysis of the DLQIMFO-RSEA approach showed superior accuracy of 94.34 % compared to other models on the student engagement dataset.

量子计算（QC）利用量子力学原理，如叠加和纠缠来解决特定问题，比经典计算机更有效。电子学习是大流行期间学生和教师的唯一选择。相反，对于教师来说，在线教学时观察每个学生的参与度是一项挑战。学生们在这样的活动中会分心。教师想要识别学生的状态，无论他们是专注于教学还是专注于教学。因此，从教师的角度来看，评估学生的参与水平以了解他们的实际反应并采取必要的步骤让学生参与并帮助他们实现目标是很重要的，机器学习（ML）和深度学习（DL）被用于根据互动、课堂贡献等对学生的表现和参与进行预测分析。本文提出了一种深度学习驱动的量子启发飞蛾火焰优化器实时学生参与分析（DLQIMFO-RSEA）方法。DLQIMFO-RSEA方法旨在对学生在在线课程中的参与度进行分类。为了实现这一点，DLQIMFO-RSEA方法使用YOLOv5目标检测模型，该模型具有骨干SPPF、CBS和CSPI-X。接下来，图像预处理阶段采用维纳滤波器（WF）去除噪声。对于特征提取，使用了InceptionResNetV2技术。此外，采用堆叠自编码器（SAE）进行检测。最后，采用量子启发飞蛾火焰优化器（QIMFO）模型进行参数整定，提高SAE模型的分类性能。DLQIMFO-RSEA方法的对比分析显示，与学生参与数据集上的其他模型相比，DLQIMFO-RSEA方法的准确率为94.34 %。

{"title":"Quantum-inspired deep learning optimisation for real-time student engagement analysis in virtual classrooms","authors":"Hanan T. Halawani , Shuhrah Alghamdi , Fahad Ali Aloufi , Maryam Alsolami , Abdulellah Alsulaimani , Hassan M. Aljohani","doi":"10.1016/j.aej.2025.12.019","DOIUrl":"10.1016/j.aej.2025.12.019","url":null,"abstract":"<div><div>Quantum computing (QC) employs quantum-mechanical principles such as superposition and entanglement to solve specific problems far more efficiently than classical computers. E-learning is the only option for students and teachers during the pandemic. Conversely, it is challenging for an instructor to observe every student’s engagement while educating online. Students are distracted during such activities. The teachers want to identify their students' states, whether they are concentrated or concerned with teaching. Hence, from a teacher’s viewpoint, it is significant to assess students' levels of engagement to understand their actual reactions and take the necessary steps to involve students and help them achieve goals, machine learning (ML) and deep learning (DL) is deployed for predictive analytics of a student’s performance and engagement depending upon interactions, contribution in class, etc. In this paper, the Deep Learning-Driven Quantum Inspired Moth Flame Optimizer for Real-Time Student Engagement Analysis (DLQIMFO-RSEA) method is proposed. The DLQIMFO-RSEA method aims to categorise student engagement in online classes. To accomplish this, the DLQIMFO-RSEA method uses the YOLOv5 object detection model with backbones SPPF, CBS, and CSPI-X. Next, the image pre-processing stage employs the Wiener filter (WF) to remove the noise. For feature extraction, the InceptionResNetV2 technique is used. Furthermore, a stacked autoencoder (SAE) is applied for detection. At last, the parameter tuning process is performed by the quantum-inspired moth flame optimiser (QIMFO) model to improve the classification performance of the SAE model. The comparison analysis of the DLQIMFO-RSEA approach showed superior accuracy of 94.34 % compared to other models on the student engagement dataset.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"134 ","pages":"Pages 375-390"},"PeriodicalIF":6.8,"publicationDate":"2025-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145787962","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

NumLin-Mamba-YOLO: A YOLO object detection algorithm based on Mamba architecture and multi-scale feature optimization NumLin-Mamba-YOLO：一种基于Mamba架构和多尺度特征优化的YOLO目标检测算法

IF 6.8 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

alexandria engineering journal

Pub Date : 2025-12-19 DOI: 10.1016/j.aej.2025.11.011

Qinnan Luo

Object detection in complex scenes, especially for small object detection, faces significant challenges in real-time performance and multi-scale accuracy. This study proposes an improved algorithm, NumLin-Mamba-YOLO, designed to enhance detection performance by integrating numerical linear algebra techniques with the Mamba module. The aim of this research is to optimize detection performance through a series of multi-module innovations. The algorithm utilizes Singular Value Decomposition (SVD) and Principal Component Analysis (PCA) in the backbone network for feature dimensionality reduction, denoising, and redundancy elimination, thereby enhancing feature discriminability. It also leverages the linear time complexity advantage of Mamba’s State Space Model (SSM) to model global dependencies. An enhanced feature fusion network improves cross-scale feature correlation, while a decoupled attention detection head is designed to independently optimize classification and regression tasks, improving sensitivity to small objects and local details. Experiments on the Visdrone and PASCAL VOC datasets demonstrate that the algorithm achieves substantial improvements in both detection accuracy and inference efficiency in complex and general scenes, with particular strength in small object detection. The model effectively controls parameters and computational load, providing an efficient solution for real-world object detection applications in intelligent monitoring and autonomous driving.

复杂场景下的目标检测，特别是小目标检测，在实时性和多尺度精度方面面临着巨大的挑战。本研究提出了一种改进算法NumLin-Mamba-YOLO，旨在通过将数值线性代数技术与Mamba模块相结合来提高检测性能。本研究的目的是通过一系列多模块创新来优化检测性能。该算法利用主干网络中的奇异值分解（SVD）和主成分分析（PCA）对特征进行降维、去噪和冗余消除，从而增强特征的可分辨性。它还利用了Mamba的状态空间模型（SSM）的线性时间复杂度优势来建模全局依赖关系。增强的特征融合网络提高了跨尺度特征相关性，同时设计了解耦的注意检测头，独立优化分类和回归任务，提高了对小目标和局部细节的灵敏度。在Visdrone和PASCAL VOC数据集上的实验表明，该算法在复杂场景和一般场景下的检测精度和推理效率都有了很大的提高，在小目标检测方面尤为突出。该模型有效地控制了参数和计算负荷，为智能监控和自动驾驶中的现实世界目标检测应用提供了高效的解决方案。

{"title":"NumLin-Mamba-YOLO: A YOLO object detection algorithm based on Mamba architecture and multi-scale feature optimization","authors":"Qinnan Luo","doi":"10.1016/j.aej.2025.11.011","DOIUrl":"10.1016/j.aej.2025.11.011","url":null,"abstract":"<div><div>Object detection in complex scenes, especially for small object detection, faces significant challenges in real-time performance and multi-scale accuracy. This study proposes an improved algorithm, NumLin-Mamba-YOLO, designed to enhance detection performance by integrating numerical linear algebra techniques with the Mamba module. The aim of this research is to optimize detection performance through a series of multi-module innovations. The algorithm utilizes Singular Value Decomposition (SVD) and Principal Component Analysis (PCA) in the backbone network for feature dimensionality reduction, denoising, and redundancy elimination, thereby enhancing feature discriminability. It also leverages the linear time complexity advantage of Mamba’s State Space Model (SSM) to model global dependencies. An enhanced feature fusion network improves cross-scale feature correlation, while a decoupled attention detection head is designed to independently optimize classification and regression tasks, improving sensitivity to small objects and local details. Experiments on the Visdrone and PASCAL VOC datasets demonstrate that the algorithm achieves substantial improvements in both detection accuracy and inference efficiency in complex and general scenes, with particular strength in small object detection. The model effectively controls parameters and computational load, providing an efficient solution for real-world object detection applications in intelligent monitoring and autonomous driving.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"134 ","pages":"Pages 417-432"},"PeriodicalIF":6.8,"publicationDate":"2025-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145788025","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

SVD-BDRL: A trustworthy autonomous driving decision framework based on sparse voxels and blockchain enhancement 基于稀疏体素和区块链增强的可靠自动驾驶决策框架

IF 6.8 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

alexandria engineering journal

Pub Date : 2025-12-19 DOI: 10.1016/j.aej.2025.11.040

Zhongsheng Tang , Yetao Feng , Jian Zhang , Zihao Wang

With the rapid development of autonomous driving technology, issues regarding safe decision-making and multi-vehicle collaboration in complex urban environments have become increasingly prominent. To address the limitations of existing deep reinforcement learning methods in computational efficiency, decision transparency, and system safety, this paper proposes a novel framework, SVD-BDRL, which integrates a sparse voxel decoder and blockchain-enhanced deep reinforcement learning. The framework brings three key innovations: a sparse voxelization method to reduce computational complexity at the perception layer; a blockchain-based distributed experience management system to ensure data authenticity at the decision layer; and a real-time anomaly detection system combining graph neural networks and consortium blockchain for verification. Experimental results demonstrate that on the NuScenes and CARLA datasets, SVD-BDRL outperforms current methods, achieving an 11% reduction in collision rate and a 3.4% decrease in trajectory error, while maintaining real-time performance at 23.5 FPS. This study presents a promising new approach for creating safe, trustworthy autonomous driving systems, which is crucial for the commercialization of autonomous vehicles.

随着自动驾驶技术的快速发展，复杂城市环境下的安全决策和多车协同问题日益突出。为了解决现有深度强化学习方法在计算效率、决策透明度和系统安全性方面的局限性，本文提出了一种新的框架SVD-BDRL，该框架集成了稀疏体素解码器和区块链增强的深度强化学习。该框架带来了三个关键创新：一种稀疏体素化方法，以降低感知层的计算复杂度；基于区块链的分布式体验管理系统，确保决策层数据的真实性；并结合图神经网络和财团区块链的实时异常检测系统进行验证。实验结果表明，在NuScenes和CARLA数据集上，SVD-BDRL优于现有方法，碰撞率降低11%，轨迹误差降低3.4%，同时保持23.5 FPS的实时性能。这项研究为创建安全、可靠的自动驾驶系统提供了一种有希望的新方法，这对自动驾驶汽车的商业化至关重要。

{"title":"SVD-BDRL: A trustworthy autonomous driving decision framework based on sparse voxels and blockchain enhancement","authors":"Zhongsheng Tang , Yetao Feng , Jian Zhang , Zihao Wang","doi":"10.1016/j.aej.2025.11.040","DOIUrl":"10.1016/j.aej.2025.11.040","url":null,"abstract":"<div><div>With the rapid development of autonomous driving technology, issues regarding safe decision-making and multi-vehicle collaboration in complex urban environments have become increasingly prominent. To address the limitations of existing deep reinforcement learning methods in computational efficiency, decision transparency, and system safety, this paper proposes a novel framework, SVD-BDRL, which integrates a sparse voxel decoder and blockchain-enhanced deep reinforcement learning. The framework brings three key innovations: a sparse voxelization method to reduce computational complexity at the perception layer; a blockchain-based distributed experience management system to ensure data authenticity at the decision layer; and a real-time anomaly detection system combining graph neural networks and consortium blockchain for verification. Experimental results demonstrate that on the NuScenes and CARLA datasets, SVD-BDRL outperforms current methods, achieving an 11% reduction in collision rate and a 3.4% decrease in trajectory error, while maintaining real-time performance at 23.5 FPS. This study presents a promising new approach for creating safe, trustworthy autonomous driving systems, which is crucial for the commercialization of autonomous vehicles.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"134 ","pages":"Pages 433-446"},"PeriodicalIF":6.8,"publicationDate":"2025-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145788024","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

EmoMamFusion: A sentiment classification algorithm framework based on a differentiated dual-branch fusion architecture EmoMamFusion：基于差异化双分支融合架构的情感分类算法框架

IF 6.8 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

alexandria engineering journal

Pub Date : 2025-12-19 DOI: 10.1016/j.aej.2025.11.036

Mohan Wang , Zhong Wen

Multimodal sentiment recognition enables precise analysis and classification of human emotions. However, existing methods face three major challenges: low feature fusion efficiency due to modality heterogeneity, insufficient long-range dependency modeling capability, and high model complexity. To address these issues, this paper proposes the EmoMamFusion model for multimodal sentiment recognition, based on a differentiated dual-branch fusion architecture. Specifically, the model achieves differentiated fusion through the TranFusion and MamFusion branches: TranFusion, centered around cross-attention, performs preliminary cross-modal semantic alignment and shallow feature fusion; MamFusion models long-range dependencies through the Mamba network and enforces cross-modal collaboration with InfoNCE loss, enabling deep feature fusion and semantic enhancement. Finally, the features from both branches are globally integrated by a Transformer encoder to complete sentiment category prediction. Experiments on the CMU-MOSI and CMU-MOSEI datasets show that EmoMamFusion outperforms 22 state-of-the-art methods in core metrics such as binary classification accuracy (Acc-2), F1 score, seven-class accuracy (Acc-7), and Pearson correlation coefficient (Corr). Specifically, on the CMU-MOSI dataset, Acc-7 and Corr reached 50.96% and 0.895, respectively, while on the CMU-MOSEI dataset, Acc-7 and Corr reached 54.98% and 0.821. At the same time, the model has only 1.38M parameters, achieving a collaborative optimization of high performance and lightweight design.

多模态情感识别能够对人类情感进行精确的分析和分类。然而，现有方法面临着三方面的挑战：模态异构导致特征融合效率低；远程依赖建模能力不足；模型复杂性高。为了解决这些问题，本文提出了基于差异化双分支融合架构的EmoMamFusion多模态情感识别模型。具体来说，该模型通过TranFusion和MamFusion分支实现差异化融合：TranFusion以交叉注意为中心，进行初步的跨模态语义对齐和浅层特征融合；MamFusion通过Mamba网络建立远程依赖关系模型，并通过InfoNCE loss实现跨模式协作，从而实现深度特征融合和语义增强。最后，通过Transformer编码器将来自两个分支的特征进行全局集成，以完成情感类别预测。在CMU-MOSI和CMU-MOSEI数据集上的实验表明，EmoMamFusion在二进制分类精度（Acc-2）、F1分数、七类精度（Acc-7）和Pearson相关系数（Corr）等核心指标上优于22种最先进的方法。其中，CMU-MOSI数据集的Acc-7和Corr分别达到50.96%和0.895，而CMU-MOSEI数据集的Acc-7和Corr分别达到54.98%和0.821。同时，该模型只有1.38M个参数，实现了高性能和轻量化设计的协同优化。

{"title":"EmoMamFusion: A sentiment classification algorithm framework based on a differentiated dual-branch fusion architecture","authors":"Mohan Wang , Zhong Wen","doi":"10.1016/j.aej.2025.11.036","DOIUrl":"10.1016/j.aej.2025.11.036","url":null,"abstract":"<div><div>Multimodal sentiment recognition enables precise analysis and classification of human emotions. However, existing methods face three major challenges: low feature fusion efficiency due to modality heterogeneity, insufficient long-range dependency modeling capability, and high model complexity. To address these issues, this paper proposes the EmoMamFusion model for multimodal sentiment recognition, based on a differentiated dual-branch fusion architecture. Specifically, the model achieves differentiated fusion through the TranFusion and MamFusion branches: TranFusion, centered around cross-attention, performs preliminary cross-modal semantic alignment and shallow feature fusion; MamFusion models long-range dependencies through the Mamba network and enforces cross-modal collaboration with InfoNCE loss, enabling deep feature fusion and semantic enhancement. Finally, the features from both branches are globally integrated by a Transformer encoder to complete sentiment category prediction. Experiments on the CMU-MOSI and CMU-MOSEI datasets show that EmoMamFusion outperforms 22 state-of-the-art methods in core metrics such as binary classification accuracy (Acc-2), F1 score, seven-class accuracy (Acc-7), and Pearson correlation coefficient (Corr). Specifically, on the CMU-MOSI dataset, Acc-7 and Corr reached 50.96% and 0.895, respectively, while on the CMU-MOSEI dataset, Acc-7 and Corr reached 54.98% and 0.821. At the same time, the model has only 1.38M parameters, achieving a collaborative optimization of high performance and lightweight design.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"134 ","pages":"Pages 447-459"},"PeriodicalIF":6.8,"publicationDate":"2025-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145788023","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Enhanced genetic algorithm-optimized deep learning features for lung cancer classification 增强遗传算法优化的肺癌分类深度学习特征

IF 6.8 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

alexandria engineering journal

Pub Date : 2025-12-18 DOI: 10.1016/j.aej.2025.12.023

Hanan M. Alghamdi

Lung cancer is a major global health problem and early detection is crucial to prevent serious health problems. Analyzing affected areas manually is difficult and requires skilled physicians. To help with this, AI tools have been developed for automated lung cancer detection. This paper proposes a novel method for detecting lung cancer using deep learning techniques. It uses two specific deep learning models, EfficientNet-b0 and InceptionResNet-V2, to compute important features from lung images. The EfficientNet-b0 model is adjusted by replacing certain layers to improve its performance in lung cancer data. After extracting the characteristics, a specially designed genetic algorithm helps select the most useful characteristics, reducing unnecessary features, and making the system more efficient. The improved GA reduces more than 50% of features without interfering with the recognition results. The proposed approach is validated on two publicly available datasets, the CT Scan Images for Lung Cancer Dataset and the IQ-OTH/NCCD Dataset, achieving a classification precision of 99.50% and 99.20%, respectively. Our approach achieved 0.5 to 2.5% higher accuracy in comparison to state-of-the-art methods while reducing the dimensionality of the features by more than 50%, without affecting the classification performance. The improved genetic algorithm smartly chooses key features, thus accelerating processing and lowering costs. It proves valuable in real-time medical applications and automated lung cancer detection, supporting early diagnosis and treatment planning.

肺癌是一个重大的全球健康问题，早期发现对于预防严重的健康问题至关重要。手工分析受影响的区域是困难的，需要熟练的医生。为了帮助实现这一目标，人们开发了用于肺癌自动检测的人工智能工具。本文提出了一种利用深度学习技术检测肺癌的新方法。它使用两种特定的深度学习模型，EfficientNet-b0和InceptionResNet-V2，来计算肺部图像的重要特征。通过替换某些层来调整EfficientNet-b0模型，以提高其在肺癌数据中的表现。在提取特征后，特别设计的遗传算法帮助选择最有用的特征，减少不必要的特征，提高系统的效率。改进的遗传算法在不影响识别结果的情况下，减少了50%以上的特征。该方法在CT扫描图像肺癌数据集和IQ-OTH/NCCD数据集两个公开可用的数据集上进行了验证，分类精度分别达到99.50%和99.20%。与最先进的方法相比，我们的方法的准确率提高了0.5到2.5%，同时将特征的维数降低了50%以上，而不影响分类性能。改进的遗传算法巧妙地选择关键特征，从而加快了处理速度，降低了成本。它在实时医疗应用和自动化肺癌检测中被证明是有价值的，支持早期诊断和治疗计划。

{"title":"Enhanced genetic algorithm-optimized deep learning features for lung cancer classification","authors":"Hanan M. Alghamdi","doi":"10.1016/j.aej.2025.12.023","DOIUrl":"10.1016/j.aej.2025.12.023","url":null,"abstract":"<div><div>Lung cancer is a major global health problem and early detection is crucial to prevent serious health problems. Analyzing affected areas manually is difficult and requires skilled physicians. To help with this, AI tools have been developed for automated lung cancer detection. This paper proposes a novel method for detecting lung cancer using deep learning techniques. It uses two specific deep learning models, EfficientNet-b0 and InceptionResNet-V2, to compute important features from lung images. The EfficientNet-b0 model is adjusted by replacing certain layers to improve its performance in lung cancer data. After extracting the characteristics, a specially designed genetic algorithm helps select the most useful characteristics, reducing unnecessary features, and making the system more efficient. The improved GA reduces more than 50% of features without interfering with the recognition results. The proposed approach is validated on two publicly available datasets, the CT Scan Images for Lung Cancer Dataset and the IQ-OTH/NCCD Dataset, achieving a classification precision of 99.50% and 99.20%, respectively. Our approach achieved 0.5 to 2.5% higher accuracy in comparison to state-of-the-art methods while reducing the dimensionality of the features by more than 50%, without affecting the classification performance. The improved genetic algorithm smartly chooses key features, thus accelerating processing and lowering costs. It proves valuable in real-time medical applications and automated lung cancer detection, supporting early diagnosis and treatment planning.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"134 ","pages":"Pages 341-357"},"PeriodicalIF":6.8,"publicationDate":"2025-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145787963","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Nonlinear space–time fractional KdV model for ion-acoustic shock waves in isothermal-electron, warm-ion plasmas 等温电子、暖离子等离子体中离子声激波的非线性时空分数KdV模型

IF 6.8 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

alexandria engineering journal

Pub Date : 2025-12-18 DOI: 10.1016/j.aej.2025.12.021

Zainab M.H. El-Qahtani , Fadhel Almalki , Tahani Al-Mutairi , Ali Algarni , Tahani M. Albogami , E.A.-B. Abdel-Salam

This paper investigates the nonlinear dynamics of IASW and periodic waves in warm-ion, isothermal-electron plasmas using the stfKdV equation. Exact analytical solutions are obtained through the F-expansion and Hirota’s bilinear approaches, yielding solitary, cnoidal, and multi-soliton waveforms. The effects of the fractional order α and the ion-to-electron temperature ratio σ are systematically analyzed: α regulates the effective dispersive spreading and phase evolution of the waves, while σ controls amplitude and spatial width through its influence on the nonlinear and dispersive coefficients. Phase-portrait and bifurcation analyses reveal the existence of solitary, periodic, and shock-type structures, and demonstrate how fractional-order dispersion alters the topology of the associated dynamical system. The novelties of this work lie in providing a unified analytical and dynamical framework for the stfKdV model and clarifying how α and σ jointly shape soliton morphology and interaction properties. The research enhances the theoretical framework of fractional nonlinear wave models and offers insights applicable to laboratory and space plasmas, as well as other dispersive environments where fractional-order descriptions remain effective.

本文利用stfKdV方程研究了热离子、等温电子等离子体中IASW和周期波的非线性动力学。通过f展开和Hirota的双线性方法得到了精确的解析解，得到了孤、弦、多孤子波形。系统地分析了分数阶α和离子电子温度比σ的影响：α调节波的有效色散扩展和相位演化，σ通过影响非线性系数和色散系数来控制振幅和空间宽度。相位肖像和分岔分析揭示了孤立、周期和激波型结构的存在，并展示了分数阶色散如何改变相关动力系统的拓扑结构。本工作的新颖之处在于为stfKdV模型提供了统一的分析和动力学框架，并阐明了α和σ如何共同塑造孤子的形态和相互作用性质。该研究增强了分数阶非线性波模型的理论框架，并提供了适用于实验室和空间等离子体以及分数阶描述仍然有效的其他色散环境的见解。

{"title":"Nonlinear space–time fractional KdV model for ion-acoustic shock waves in isothermal-electron, warm-ion plasmas","authors":"Zainab M.H. El-Qahtani , Fadhel Almalki , Tahani Al-Mutairi , Ali Algarni , Tahani M. Albogami , E.A.-B. Abdel-Salam","doi":"10.1016/j.aej.2025.12.021","DOIUrl":"10.1016/j.aej.2025.12.021","url":null,"abstract":"<div><div>This paper investigates the nonlinear dynamics of IASW and periodic waves in warm-ion, isothermal-electron plasmas using the stfKdV equation. Exact analytical solutions are obtained through the F-expansion and Hirota’s bilinear approaches, yielding solitary, cnoidal, and multi-soliton waveforms. The effects of the fractional order α and the ion-to-electron temperature ratio σ are systematically analyzed: α regulates the effective dispersive spreading and phase evolution of the waves, while σ controls amplitude and spatial width through its influence on the nonlinear and dispersive coefficients. Phase-portrait and bifurcation analyses reveal the existence of solitary, periodic, and shock-type structures, and demonstrate how fractional-order dispersion alters the topology of the associated dynamical system. The novelties of this work lie in providing a unified analytical and dynamical framework for the stfKdV model and clarifying how α and σ jointly shape soliton morphology and interaction properties. The research enhances the theoretical framework of fractional nonlinear wave models and offers insights applicable to laboratory and space plasmas, as well as other dispersive environments where fractional-order descriptions remain effective.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"134 ","pages":"Pages 358-374"},"PeriodicalIF":6.8,"publicationDate":"2025-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145787964","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Learnable graph ODE networks for anomaly detection in CAN-FD vehicle networks 用于CAN-FD车辆网络异常检测的可学习图ODE网络

IF 6.8 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

alexandria engineering journal

Pub Date : 2025-12-17 DOI: 10.1016/j.aej.2025.12.020

Miao Xu , Lizeng Zhang , Peiyu Hou

As vehicles become increasingly interconnected with networks, not only does communication between onboard systems grow more frequent, but real-time interactions between these systems and external infrastructure also significantly increase. Although these complex interactions promote the development of intelligent transportation, they also greatly increase the attack surface. In the face of such a background, identifying abnormal vehicle behavior in time becomes crucial to driving safety. The current anomaly detection methods cannot fully reflect the semantic information in the vehicle data. Without the knowledge of the logical relation of messages, models are very likely to produce a large amount of false positives, and it is difficult to explain the decision-making basis. To alleviate the above issues, we propose a learnable graph-based differential equation anomaly detection framework (LGOVAD) for CAN-FD vehicle networks. LGOVAD consists of three fundamental modules, which work cooperatively to interpretable and model the semantic dynamics of vehicle communication data. The Semantic Extractor decodes interpretable physical signals based on the raw CAN messages. Then the Semantic Relationship Perception module discovers the hidden dependencies between vehicle parameters dynamically by employing a dual-branch graph learning strategy. Finally, we use a graph ODE network to capture the evolution process of these semantic patterns over time in a continuous manner. Experiments on real CAN-FD datasets demonstrate LGOVAD’s ability to accurately identify semantic and temporal anomalies triggered by attack messages. Additionally, we validate the model’s cross-domain generalization capabilities on SWaT and WADI datasets. Specifically, on CAN-FD data, LGOVAD achieves F1 scores 1.27-11.51% higher than specialized in-vehicle detectors (CANet, CANShield, PSEAD) and 3.48-16.42% higher than general multivariate temporal methods (GTA, FuGLAD). while maintaining robust advantages in normalized F1 scores on industrial benchmarks (SWaT, WADI).

随着车辆越来越多地与网络互联，不仅车载系统之间的通信变得更加频繁，而且这些系统与外部基础设施之间的实时交互也大大增加。这些复杂的相互作用虽然促进了智能交通的发展，但也大大增加了攻击面。在这样的背景下，及时识别车辆异常行为对行车安全至关重要。现有的异常检测方法不能充分反映车辆数据中的语义信息。如果不了解消息的逻辑关系，模型很可能产生大量的误报，并且很难解释决策基础。为了缓解上述问题，我们提出了一种可学习的基于图的CAN-FD车辆网络微分方程异常检测框架（LGOVAD）。LGOVAD由三个基本模块组成，它们协同工作，对车辆通信数据的语义动态进行解释和建模。语义提取器基于原始CAN消息解码可解释的物理信号。语义关系感知模块采用双分支图学习策略，动态发现车辆参数之间隐藏的依赖关系。最后，我们使用图形ODE网络以连续的方式捕获这些语义模式随时间的演变过程。在真实CAN-FD数据集上的实验表明，LGOVAD能够准确识别由攻击消息触发的语义和时间异常。此外，我们在SWaT和WADI数据集上验证了模型的跨域泛化能力。具体而言，在CAN-FD数据上，LGOVAD的F1得分比专用车载检测器（CANet、CANShield、PSEAD）高1.27-11.51%，比通用多元时间方法（GTA、FuGLAD）高3.48-16.42%。同时在工业基准（SWaT， WADI）上保持标准化F1分数的强大优势。

{"title":"Learnable graph ODE networks for anomaly detection in CAN-FD vehicle networks","authors":"Miao Xu , Lizeng Zhang , Peiyu Hou","doi":"10.1016/j.aej.2025.12.020","DOIUrl":"10.1016/j.aej.2025.12.020","url":null,"abstract":"<div><div>As vehicles become increasingly interconnected with networks, not only does communication between onboard systems grow more frequent, but real-time interactions between these systems and external infrastructure also significantly increase. Although these complex interactions promote the development of intelligent transportation, they also greatly increase the attack surface. In the face of such a background, identifying abnormal vehicle behavior in time becomes crucial to driving safety. The current anomaly detection methods cannot fully reflect the semantic information in the vehicle data. Without the knowledge of the logical relation of messages, models are very likely to produce a large amount of false positives, and it is difficult to explain the decision-making basis. To alleviate the above issues, we propose a learnable graph-based differential equation anomaly detection framework (LGOVAD) for CAN-FD vehicle networks. LGOVAD consists of three fundamental modules, which work cooperatively to interpretable and model the semantic dynamics of vehicle communication data. The Semantic Extractor decodes interpretable physical signals based on the raw CAN messages. Then the Semantic Relationship Perception module discovers the hidden dependencies between vehicle parameters dynamically by employing a dual-branch graph learning strategy. Finally, we use a graph ODE network to capture the evolution process of these semantic patterns over time in a continuous manner. Experiments on real CAN-FD datasets demonstrate LGOVAD’s ability to accurately identify semantic and temporal anomalies triggered by attack messages. Additionally, we validate the model’s cross-domain generalization capabilities on SWaT and WADI datasets. Specifically, on CAN-FD data, LGOVAD achieves F1 scores 1.27-11.51% higher than specialized in-vehicle detectors (CANet, CANShield, PSEAD) and 3.48-16.42% higher than general multivariate temporal methods (GTA, FuGLAD). while maintaining robust advantages in normalized F1 scores on industrial benchmarks (SWaT, WADI).</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"134 ","pages":"Pages 308-320"},"PeriodicalIF":6.8,"publicationDate":"2025-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145788019","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

An explainable path reasoning framework for knowledge graph completion 知识图谱补全的可解释路径推理框架

IF 6.8 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

alexandria engineering journal

Pub Date : 2025-12-17 DOI: 10.1016/j.aej.2025.12.024

Haiyang Yu , Hong Yu , Wei Wang , Kai Fang , Xiaotong Zhang , Han Liu

Knowledge Graphs (KGs) are vital for structured knowledge but suffer from intrinsic incompleteness. Knowledge Graph Completion (KGC) faces a critical trade-off between the performance of ”black-box” models and the interpretability of explainable approaches that rely on purely symbolic structures. To resolve this, we propose the Explainable Path Reasoning (EPR) framework, which synergizes statistical topology analysis with deep semantic modeling. EPR first mines a corpus of statistically-grounded reasoning paths using a hop-normalized conditional likelihood to mitigate length bias. This corpus is then used to train a powerful BERT encoder via a path-level contrastive objective, teaching it to comprehend compositional semantics. For inference, a single, unified BERT architecture powers a dynamic beam search, performing the dual role of scoring candidates for prediction and constructing faithful, multi-hop explanations. Experimental results demonstrate that EPR establishes a new competitive performance among explainable methods and significantly narrows the performance gap to leading black-box models, providing a powerful and transparent solution for KGC, which is essential for trustworthy decision-making in intelligent communication systems.

知识图（KGs）是结构化知识的重要组成部分，但存在固有的不完整性问题。知识图谱补全（Knowledge Graph Completion， KGC）在“黑盒”模型的性能和依赖于纯粹符号结构的可解释方法的可解释性之间面临着一个关键的权衡。为了解决这个问题，我们提出了可解释路径推理（EPR）框架，该框架将统计拓扑分析与深度语义建模相结合。EPR首先使用跳跃归一化条件似然挖掘基于统计的推理路径语料库，以减轻长度偏差。然后使用该语料库通过路径级对比目标来训练强大的BERT编码器，教它理解组合语义。对于推理，一个单一的、统一的BERT架构支持动态波束搜索，执行双重角色，为预测评分候选并构建忠实的、多跳的解释。实验结果表明，EPR在可解释方法中建立了一种新的竞争性能，显著缩小了与领先黑箱模型的性能差距，为KGC提供了一个强大而透明的解决方案，这对智能通信系统中可信决策至关重要。

{"title":"An explainable path reasoning framework for knowledge graph completion","authors":"Haiyang Yu , Hong Yu , Wei Wang , Kai Fang , Xiaotong Zhang , Han Liu","doi":"10.1016/j.aej.2025.12.024","DOIUrl":"10.1016/j.aej.2025.12.024","url":null,"abstract":"<div><div>Knowledge Graphs (KGs) are vital for structured knowledge but suffer from intrinsic incompleteness. Knowledge Graph Completion (KGC) faces a critical trade-off between the performance of ”black-box” models and the interpretability of explainable approaches that rely on purely symbolic structures. To resolve this, we propose the Explainable Path Reasoning (EPR) framework, which synergizes statistical topology analysis with deep semantic modeling. EPR first mines a corpus of statistically-grounded reasoning paths using a hop-normalized conditional likelihood to mitigate length bias. This corpus is then used to train a powerful BERT encoder via a path-level contrastive objective, teaching it to comprehend compositional semantics. For inference, a single, unified BERT architecture powers a dynamic beam search, performing the dual role of scoring candidates for prediction and constructing faithful, multi-hop explanations. Experimental results demonstrate that EPR establishes a new competitive performance among explainable methods and significantly narrows the performance gap to leading black-box models, providing a powerful and transparent solution for KGC, which is essential for trustworthy decision-making in intelligent communication systems.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"134 ","pages":"Pages 321-328"},"PeriodicalIF":6.8,"publicationDate":"2025-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145788020","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Optimized spectral feature selection with lightweight CNN for precise volume fraction estimation in three-phase flows 优化光谱特征选择与轻量级CNN在三相流的精确体积分数估计

IF 6.8 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

alexandria engineering journal

Pub Date : 2025-12-17 DOI: 10.1016/j.aej.2025.12.009

Abdulilah Mohammad Mayet

The ability to accurately calculate volumetric fractions for three-phase flow: gas, water and oil, is critical for industrial applications in the oil and gas and chemical engineering industry. This paper develops a methodology that incorporates the acquisition of signals using X-rays, spectral features and machine learning. Spectral features representing mean energy, variance, full width at half maximum (FWHM), entropy, kurtosis, and cepstral coefficients, were obtained from 108 Monte Carlo N-Particle (MCNP) energy simulation sets through annular, homogeneous and stratified flow regimes and 36 combinations of volume fractions ranging from 10 % to 80 %. The Beetle Antennae Search (BAS) algorithm optimizes the spectral features using measures that reduce the number of features to six features to improve overall prediction efficiency, these six features are the final input for a lightweight convolutional neural network (CNN). The chosen samples from the MCNP simulations (86 samples for training input and 22 for testing input, referred to as "the dataset") were used for training and testing the model in order to predict the three volume fractions. The predictions resulted in low root mean square error (RMSE) values, documented as 0.15 for gas, 0.13 for water and 0.14 for oil with R² values all exceeding 0.94; residuals were assessed to determine the model's stability and showed that the residuals were clustered around 0 for both training and testing sets. This approach offers a real-time and computationally-efficient technique for use in industry, but the reliance on simulated data and number of existing samples means there is still room for improvement in the future. The study demonstrates a scalable method for multiphase flow analysis, which could be verified with real experimental data in future studies to improve on the overall detection method reliability.

能够准确计算三相流（气、水和油）的体积分数，对于油气和化学工程行业的工业应用至关重要。本文开发了一种方法，该方法结合了使用x射线，光谱特征和机器学习的信号采集。从108个蒙特卡罗n粒子（MCNP）能量模拟集中，通过环形、均匀和分层流型以及体积分数范围为10 %至80 %的36种组合，获得了代表平均能量、方差、半最大全宽（FWHM）、熵、峰度和倒谱系数的光谱特征。甲虫天线搜索（BAS）算法通过将特征数量减少到6个特征的措施来优化频谱特征，以提高整体预测效率，这6个特征是轻量级卷积神经网络（CNN）的最终输入。从MCNP模拟中选择的样本（86个样本用于训练输入，22个样本用于测试输入，称为“数据集”）用于训练和测试模型，以预测三个体积分数。预测结果的均方根误差（RMSE）值较低，天然气为0.15，水为0.13，油为0.14，R²值均超过0.94；对残差进行评估以确定模型的稳定性，结果表明，对于训练集和测试集，残差都聚集在0附近。这种方法为工业应用提供了一种实时且计算效率高的技术，但对模拟数据和现有样本数量的依赖意味着未来仍有改进的空间。该研究为多相流分析提供了一种可扩展的方法，可在未来的研究中用真实的实验数据进行验证，以提高整体检测方法的可靠性。

{"title":"Optimized spectral feature selection with lightweight CNN for precise volume fraction estimation in three-phase flows","authors":"Abdulilah Mohammad Mayet","doi":"10.1016/j.aej.2025.12.009","DOIUrl":"10.1016/j.aej.2025.12.009","url":null,"abstract":"<div><div>The ability to accurately calculate volumetric fractions for three-phase flow: gas, water and oil, is critical for industrial applications in the oil and gas and chemical engineering industry. This paper develops a methodology that incorporates the acquisition of signals using X-rays, spectral features and machine learning. Spectral features representing mean energy, variance, full width at half maximum (FWHM), entropy, kurtosis, and cepstral coefficients, were obtained from 108 Monte Carlo N-Particle (MCNP) energy simulation sets through annular, homogeneous and stratified flow regimes and 36 combinations of volume fractions ranging from 10 % to 80 %. The Beetle Antennae Search (BAS) algorithm optimizes the spectral features using measures that reduce the number of features to six features to improve overall prediction efficiency, these six features are the final input for a lightweight convolutional neural network (CNN). The chosen samples from the MCNP simulations (86 samples for training input and 22 for testing input, referred to as \"the dataset\") were used for training and testing the model in order to predict the three volume fractions. The predictions resulted in low root mean square error (RMSE) values, documented as 0.15 for gas, 0.13 for water and 0.14 for oil with R² values all exceeding 0.94; residuals were assessed to determine the model's stability and showed that the residuals were clustered around 0 for both training and testing sets. This approach offers a real-time and computationally-efficient technique for use in industry, but the reliance on simulated data and number of existing samples means there is still room for improvement in the future. The study demonstrates a scalable method for multiphase flow analysis, which could be verified with real experimental data in future studies to improve on the overall detection method reliability.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"134 ","pages":"Pages 329-340"},"PeriodicalIF":6.8,"publicationDate":"2025-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145788027","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

The application development of nanoparticle-based drug delivery systems in combating drug resistance of lung cancer 纳米颗粒给药系统在抗肺癌耐药中的应用进展

IF 6.8 2区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY

alexandria engineering journal

Pub Date : 2025-12-15 DOI: 10.1016/j.aej.2025.12.002

Dong Li , Xiaoqing Wang , Jinglan Zhao , Meina Shi , Baoyi Chen , Guifang Su

Lung cancer remains a leading cause of cancer mortality, with drug resistance contributing to over 90 % of therapeutic failures in advanced cases. This review systematically examines how engineered nanoparticles can address the multifaceted mechanisms of resistance—including genetic mutations, efflux transporter overexpression, tumor microenvironment alterations, and stemness-associated phenotypes. Nanocarriers such as liposomes, polymeric nanoparticles, solid lipid nanoparticles, dendrimers, and inorganic particles have demonstrated enhanced tumor accumulation, prolonged circulation, and triggered intracellular release. In multidrug-resistant models, these formulations reduced doxorubicin IC50 values by up to 90 % and restored sensitivity to paclitaxel and cisplatin. Key strategies include co-delivering chemotherapeutics with agents like siRNA or modulators, achieving significant tumor shrinkage and re-sensitization, and employing stimuli-responsive designs for precise payload release at tumor sites, thereby reducing systemic toxicity. Organelle-targeted delivery (mitochondria, nuclei) further enhances efficacy against resistant subpopulations. In preclinical settings, these strategies have not only achieved tumor regression but also reduced metastases and improved overall survival. The integration of active targeting ligands and immunomodulatory agents positions nanoparticle platforms as multifunctional systems with substantial translational potential. This review highlights the engineering innovations that enable precise delivery, effective drug release, and circumvention of resistance in lung cancer models.

肺癌仍然是癌症死亡的主要原因，在晚期病例中，90%以上的治疗失败都是耐药性造成的 %。这篇综述系统地研究了工程纳米颗粒如何解决耐药的多方面机制，包括基因突变、外排转运体过表达、肿瘤微环境改变和干细胞相关表型。纳米载体如脂质体、聚合纳米颗粒、固体脂质纳米颗粒、树状大分子和无机颗粒已被证明可以增强肿瘤积累、延长循环并触发细胞内释放。在多药耐药模型中，这些制剂将阿霉素IC50值降低了90% %，并恢复了对紫杉醇和顺铂的敏感性。关键策略包括与siRNA或调节剂等药物共同递送化疗药物，实现显著的肿瘤缩小和再致敏，以及采用刺激响应设计在肿瘤部位精确释放有效载荷，从而降低全身毒性。细胞器靶向递送（线粒体、细胞核）进一步增强了对耐药亚群的疗效。在临床前，这些策略不仅实现了肿瘤消退，还减少了转移，提高了总生存率。主动靶向配体和免疫调节剂的整合使纳米颗粒平台成为具有巨大翻译潜力的多功能系统。这篇综述强调了在肺癌模型中实现精确递送、有效药物释放和规避耐药的工程创新。

{"title":"The application development of nanoparticle-based drug delivery systems in combating drug resistance of lung cancer","authors":"Dong Li , Xiaoqing Wang , Jinglan Zhao , Meina Shi , Baoyi Chen , Guifang Su","doi":"10.1016/j.aej.2025.12.002","DOIUrl":"10.1016/j.aej.2025.12.002","url":null,"abstract":"<div><div>Lung cancer remains a leading cause of cancer mortality, with drug resistance contributing to over 90 % of therapeutic failures in advanced cases. This review systematically examines how engineered nanoparticles can address the multifaceted mechanisms of resistance—including genetic mutations, efflux transporter overexpression, tumor microenvironment alterations, and stemness-associated phenotypes. Nanocarriers such as liposomes, polymeric nanoparticles, solid lipid nanoparticles, dendrimers, and inorganic particles have demonstrated enhanced tumor accumulation, prolonged circulation, and triggered intracellular release. In multidrug-resistant models, these formulations reduced doxorubicin IC50 values by up to 90 % and restored sensitivity to paclitaxel and cisplatin. Key strategies include co-delivering chemotherapeutics with agents like siRNA or modulators, achieving significant tumor shrinkage and re-sensitization, and employing stimuli-responsive designs for precise payload release at tumor sites, thereby reducing systemic toxicity. Organelle-targeted delivery (mitochondria, nuclei) further enhances efficacy against resistant subpopulations. In preclinical settings, these strategies have not only achieved tumor regression but also reduced metastases and improved overall survival. The integration of active targeting ligands and immunomodulatory agents positions nanoparticle platforms as multifunctional systems with substantial translational potential. This review highlights the engineering innovations that enable precise delivery, effective drug release, and circumvention of resistance in lung cancer models.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"134 ","pages":"Pages 275-307"},"PeriodicalIF":6.8,"publicationDate":"2025-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145788021","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0