Artificial Intelligence Review最新文献

{"title":"A survey of small object detection based on deep learning in aerial images","authors":"Wei Hua,&nbsp;Qili Chen","doi":"10.1007/s10462-025-11150-9","DOIUrl":"10.1007/s10462-025-11150-9","url":null,"abstract":"<div><p>Small object detection poses a formidable challenge in the field of computer vision, particularly when it comes to analyzing aerial remote sensing images. Despite the rapid development of deep learning and significant progress in detection techniques in natural scenes, the migration of these algorithms to aerial images has not met expectations. This is primarily due to limitations in imaging acquisition conditions, including small target size, viewpoint specificity, background complexity, as well as scale and orientation diversity. Although the increasing application of deep learning-based algorithms to overcome these problems, few studies have summarized the optimization of different deep learning strategies used for small target detection in aerial images. Therefore, this paper aims to explore the application of deep learning methods for small object detection in aerial images. The primary challenges in small object detection in aerial images will be summarized. Next, a meticulous analysis and categorization of the prevailing deep learning optimization strategies employed to surmount the challenges encountered in aerial image detection is undertaken. Following that, we provide a comprehensive presentation of the object detection datasets utilized in aerial remote sensing images, along with the evaluation metrics employed. Additionally, we furnish experimental data pertaining to the currently proposed detection algorithms. Finally, the advantages and disadvantages of various optimization strategies and potential development trends are discussed. Hopefully, it can provide a reference for researchers in this field.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"58 6","pages":""},"PeriodicalIF":10.7,"publicationDate":"2025-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-025-11150-9.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143621803","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"BERT applications in natural language processing: a review","authors":"Nadia Mushtaq Gardazi,&nbsp;Ali Daud,&nbsp;Muhammad Kamran Malik,&nbsp;Amal Bukhari,&nbsp;Tariq Alsahfi,&nbsp;Bader Alshemaimri","doi":"10.1007/s10462-025-11162-5","DOIUrl":"10.1007/s10462-025-11162-5","url":null,"abstract":"<div><p>BERT (Bidirectional Encoder Representations from Transformers) has revolutionized Natural Language Processing (NLP) by significantly enhancing the capabilities of language models. This review study examines the complex nature of BERT, including its structure, utilization in different NLP tasks, and the further development of its design via modifications. The study thoroughly analyses the methodological aspects, conducting a comprehensive analysis of the planning process, the implemented procedures, and the criteria used to decide which data to include or exclude in the evaluation framework. In addition, the study thoroughly examines the influence of BERT on several NLP tasks, such as Sentence Boundary Detection, Tokenization, Grammatical Error Detection and Correction, Dependency Parsing, Named Entity Recognition, Part of Speech Tagging, Question Answering Systems, Machine Translation, Sentiment analysis, fake review detection and Cross-lingual transfer learning. The review study adds to the current literature by integrating ideas from multiple sources, explicitly emphasizing the problems and prospects in BERT-based models. The objective is to comprehensively comprehend BERT and its implementations, targeting both experienced researchers and novices in the domain of NLP. Consequently, the present study is expected to inspire more research endeavors, promote innovative adaptations of BERT, and deepen comprehension of its extensive capabilities in various NLP applications. The results presented in this research are anticipated to influence the advancement of future language models and add to the ongoing discourse on enhancing technology for understanding natural language.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"58 6","pages":""},"PeriodicalIF":10.7,"publicationDate":"2025-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-025-11162-5.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143621806","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Intelligent games meeting with multi-agent deep reinforcement learning: a comprehensive review","authors":"Yiqin Wang,&nbsp;Yufeng Wang,&nbsp;Feng Tian,&nbsp;Jianhua Ma,&nbsp;Qun Jin","doi":"10.1007/s10462-025-11166-1","DOIUrl":"10.1007/s10462-025-11166-1","url":null,"abstract":"<div><p>Recent years have witnessed the great achievement of the AI-driven intelligent games, such as AlphaStar defeating the human experts, and numerous intelligent games have come into the public view. Essentially, deep reinforcement learning (DRL), especially multiple-agent DRL (MADRL) has empowered a variety of artificial intelligence fields, including intelligent games. However, there is lack of systematical review on their correlations. This article provides a holistic picture on smoothly connecting intelligent games with MADRL from two perspectives: theoretical game concepts for MADRL, and MADRL for intelligent games. From the first perspective, information structure and game environmental features for MADRL algorithms are summarized; and from the second viewpoint, the challenges in intelligent games are investigated, and the existing MADRL solutions are correspondingly explored. Furthermore, the state-of-the-art (SOTA) MADRL algorithms for intelligent games are systematically categorized, especially from the perspective of credit assignment. Moreover, a comprehensively review on notorious benchmarks are conducted to facilitate the design and test of MADRL based intelligent games. Besides, a general procedure of MADRL simulations is offered. Finally, the key challenges in integrating intelligent games with MADRL, and potential future research directions are highlighted. This survey hopes to provide a thoughtful insight of developing intelligent games with the assistance of MADRL solutions and algorithms.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"58 6","pages":""},"PeriodicalIF":10.7,"publicationDate":"2025-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-025-11166-1.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143621800","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"GFSNet: Gaussian Fourier with sparse attention network for visual question answering","authors":"Xiang Shen,&nbsp;Dezhi Han,&nbsp;Chin-Chen Chang,&nbsp;Ammar Oad,&nbsp;Huafeng Wu","doi":"10.1007/s10462-025-11163-4","DOIUrl":"10.1007/s10462-025-11163-4","url":null,"abstract":"<div><p>Visual question answering (VQA), a core task in multimodal learning, requires models to effectively integrate visual and natural language information to perform reasoning and semantic understanding in complex scenarios. However, self-attention mechanisms often struggle to capture multi-scale information and key region features within images comprehensively. Moreover, VQA involves multidimensional and deep reasoning about image content, particularly in scenarios involving spatial relationships and frequency-domain features. Existing methods face limitations in modeling multi-scale features and filtering irrelevant information effectively. This paper proposes an innovative Gaussian Fourier with Sparse Attention Network (GFSNet) to address these challenges. GFSNet leverages Fourier transforms to map image attention weights generated by the self-attention mechanism from the spatial domain to the frequency domain, enabling comprehensive modeling of multi-scale frequency information. This enhances the model’s adaptability to complex structures and its capacity for relational modeling. To further improve feature robustness, a Gaussian filter is introduced to suppress high-frequency noise in the frequency domain, preserving critical visual information. Additionally, a sparse attention mechanism dynamically selects optimized frequency-domain features, effectively reducing interference from redundant information while improving interpretability and computational efficiency. Without increasing parameter counts or computational complexity, GFSNet achieves efficient modeling of multi-scale visual information. Experimental results on benchmark VQA datasets (VQA v2, GQA, and CLEVR) demonstrate that GFSNet significantly enhances reasoning capabilities and cross-modal alignment performance, validating its superiority and effectiveness. The code is available at https://github.com/shenxiang-vqa/GFSNet.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"58 6","pages":""},"PeriodicalIF":10.7,"publicationDate":"2025-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-025-11163-4.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143621808","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Adaptive gaining-sharing knowledge-based variant algorithm with historical probability expansion and its application in escape maneuver decision making","authors":"Lei Xie,&nbsp;Yuan Wang,&nbsp;Shangqin Tang,&nbsp;Yintong Li,&nbsp;Zhuoran Zhang,&nbsp;Changqiang Huang","doi":"10.1007/s10462-024-11096-4","DOIUrl":"10.1007/s10462-024-11096-4","url":null,"abstract":"<div><p>To further improve the performance of adaptive gaining-sharing knowledge-based algorithm (AGSK), a novel adaptive gaining sharing knowledge-based algorithm with historical probability expansion (HPE-AGSK) is proposed by modifying the search strategies. Based on AGSK, three improvement strategies are proposed. First, expansion sharing strategy is proposed and added in junior gaining-sharing phase to boost local search ability. Second, historical probability expansion strategy is proposed and added in senior gaining-sharing phase to strengthen global search ability. Last, reverse gaining strategy is proposed and utilized to expand population distribution at the beginning of iterations. The performance of HPE-AGSK is initially evaluated using IEEE CEC 2021 test suite, compared with fifteen state-of-the-art algorithms (AGSK, APGSK, APGSK-IMODE, GLAGSK, EDA2, AAVS-EDA, EBOwithCMAR, LSHADE-SPACMA, HSES, IMODE, MadDE, CJADE, and iLSHADE-RSP). The results demonstrate that HPE-AGSK outperforms both state-of-the-art GSK-based variants and past winners of IEEE CEC competitions. Subsequently, GSK-based variants and other exceptional algorithms in CEC 2021 are selected to further evaluate the performance of HPE-AGSK using IEEE CEC 2018 test suite. The statistical results show that HPE-AGSK has superior exploration ability than the comparison algorithms, and has strong competition with APGSK (state-of-the-art AGSK variant) and IMODE (CEC 2020 Winner) in exploitation ability. Finally, HPE-AGSK is utilized to solve the beyond visual range escape maneuver decision making problem. Its success rate is 100%, and mean maneuver time is 9.10 s, these results show that HPE-AGSK has good BVR escape maneuver decision-making performance. In conclusion, HPE-AGSK is a highly promising AGSK variant that significantly enhances the performance, and is an outstanding development of AGSK. The code of HPE-AGSK can be downloaded from https://github.com/xieleilei0305/HPE-AGSK-CODE.git. (The link will be available for readers after the paper is published).</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"58 6","pages":""},"PeriodicalIF":10.7,"publicationDate":"2025-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-024-11096-4.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143621807","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Learn to optimise for job shop scheduling: a survey with comparison between genetic programming and reinforcement learning","authors":"Meng Xu,&nbsp;Yi Mei,&nbsp;Fangfang Zhang,&nbsp;Mengjie Zhang","doi":"10.1007/s10462-024-11059-9","DOIUrl":"10.1007/s10462-024-11059-9","url":null,"abstract":"<div><p>Job shop scheduling holds significant importance due to its relevance and impact on various industrial and manufacturing processes. It involves dynamically assigning and sequencing jobs to machines in a flexible production environment, where job characteristics, machine availability, and other factors might change over time. Genetic programming and reinforcement learning have emerged as powerful approaches to automatically learn high-quality scheduling heuristics or directly optimise sequences of specific job-machine pairs to generate efficient schedules in manufacturing. Existing surveys on job shop scheduling typically provide overviews from a singular perspective, focusing solely on genetic programming or reinforcement learning, but overlook the hybridisation and comparison of both approaches. This survey aims to bridge this gap by reviewing recent developments in genetic programming and reinforcement learning approaches for job shop scheduling problems, providing a comparison in terms of the learning principles and characteristics for solving different kinds of job shop scheduling problems. In addition, this survey identifies and discusses current issues and challenges in the field of learning to optimise for job shop scheduling. This comprehensive exploration of genetic programming and reinforcement learning in job shop scheduling provides valuable insights into the learning principles for optimising different job shop scheduling problems. It deepens our understanding of recent developments, suggesting potential research directions for future advancements.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"58 6","pages":""},"PeriodicalIF":10.7,"publicationDate":"2025-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-024-11059-9.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143621991","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Environmental sound recognition on embedded devices using deep learning: a review","authors":"Pau Gairí,&nbsp;Tomàs Pallejà,&nbsp;Marcel Tresanchez","doi":"10.1007/s10462-025-11106-z","DOIUrl":"10.1007/s10462-025-11106-z","url":null,"abstract":"<div><p>Sound recognition has a wide range of applications beyond speech and music, including environmental monitoring, sound source classification, mechanical fault diagnosis, audio fingerprinting, and event detection. These applications often require real-time data processing, making them well-suited for embedded systems. However, embedded devices face significant challenges due to limited computational power, memory, and low power consumption. Despite these constraints, achieving high performance in environmental sound recognition typically requires complex algorithms. Deep Learning models have demonstrated high accuracy on existing datasets, making them a popular choice for such tasks. However, these models are resource-intensive, posing challenges for real-time edge applications. This paper presents a comprehensive review of integrating Deep Learning models into embedded systems, examining their state-of-the-art applications, key components, and steps involved. It also explores strategies to optimise performance in resource-constrained environments through a comparison of various implementation approaches such as knowledge distillation, pruning, and quantization, with studies achieving a reduction in complexity of up to 97% compared to the unoptimized model. Overall, we conclude that in spite of the availability of lightweight deep learning models, input features, and compression techniques, their integration into low-resource devices, such as microcontrollers, remains limited. Furthermore, more complex tasks, such as general sound classification, especially with expanded frequency bands and real-time operation have yet to be effectively implemented on these devices. These findings highlight the need for a standardised research framework to evaluate these technologies applied to resource-constrained devices, and for further development to realise the wide range of potential applications.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"58 6","pages":""},"PeriodicalIF":10.7,"publicationDate":"2025-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-025-11106-z.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143621801","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Quantum encoding whale optimization algorithm for global optimization and adaptive infinite impulse response system identification","authors":"Jinzhong Zhang,&nbsp;Wei Liu,&nbsp;Gang Zhang,&nbsp;Tan Zhang","doi":"10.1007/s10462-025-11120-1","DOIUrl":"10.1007/s10462-025-11120-1","url":null,"abstract":"<div><p>The whale optimization algorithm (WOA) is motivated by the predatory nature of bubble nets and mimics dwindling and encircling, bubble net persecuting, and randomized wandering and foraging actions to locate the expansive adequate value. However, the WOA has several deficiencies: inadequate resolution accuracy, sluggish convergence speed, susceptibility to search stagnation, and insufficient localized detection efficiency. A quantum encoding WOA (QWOA) is introduced for global optimization and adaptive infinite impulse response (IIR) system identification. The quantum encoding mechanism exploits the principle of a quantum bit to encode a search agent, which manipulates the state of an essential quantum bit and amends the location data. The quantum rotation gate modulates the quantum bit’s configuration, the quantum NOT gate accomplishes bit mutation and prohibits precocious convergence. The probability amplitude of the quantum bit represents the multistate superposition state of the search agent, which enriches the population diversity, advances individualized information, broadens the detection scope, inhibits premature convergence, facilitates estimation effectiveness, and promotes solution accuracy. The QWOA not only promptly locates the search scope nearest the most appropriate solution but also computes the spiral-shaped encircling route to promote predation diversification. Twenty-three benchmark functions, eight real-world engineering layouts, and adaptive IIR system identification are utilized to assess the QWOA’s feasibility and effectiveness. The experimental results reveal that QWOA successfully equalizes exploration and exploitation to accelerate convergence speed, ameliorate calculation accuracy, and strengthen stability and robustness.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"58 5","pages":""},"PeriodicalIF":10.7,"publicationDate":"2025-03-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-025-11120-1.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143581300","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Semi-supervised constrained clustering: an in-depth overview, ranked taxonomy and future research directions","authors":"Germán González-Almagro,&nbsp;Daniel Peralta,&nbsp;Eli De Poorter,&nbsp;José-Ramón Cano,&nbsp;Salvador García","doi":"10.1007/s10462-024-11103-8","DOIUrl":"10.1007/s10462-024-11103-8","url":null,"abstract":"<div><p>Clustering is a well-known unsupervised machine learning approach capable of automatically grouping discrete sets of instances with similar characteristics. Constrained clustering is a semi-supervised extension to this process that can be used when expert knowledge is available to indicate constraints that can be exploited. Well-known examples of such constraints are must-link (indicating that two instances belong to the same group) and cannot-link (two instances definitely do not belong together). The research area of constrained clustering has grown significantly over the years with a large variety of new algorithms and more advanced types of constraints being proposed. However, no unifying overview is available to easily understand the wide variety of available methods, constraints and benchmarks. To remedy this, this study presents in-detail the background of constrained clustering and provides a novel ranked taxonomy of the types of constraints that can be used in constrained clustering. In addition, it focuses on the instance-level pairwise constraints, and gives an overview of its applications and its historical context. Finally, it presents a statistical analysis covering 315 constrained clustering methods, categorizes them according to their features, and provides a ranking score indicating which methods have the most potential based on their popularity and validation quality. Finally, based upon this analysis, potential pitfalls and future research directions are provided.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"58 5","pages":""},"PeriodicalIF":10.7,"publicationDate":"2025-03-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-024-11103-8.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143564508","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Bridging machine learning and peptide design for cancer treatment: a comprehensive review","authors":"Khosro Rezaee,&nbsp;Hossein Eslami","doi":"10.1007/s10462-025-11148-3","DOIUrl":"10.1007/s10462-025-11148-3","url":null,"abstract":"<div><p>Anticancer peptides (ACPs) offer a promising alternative to traditional cancer therapies due to their specificity and reduced side effects. The development of ACPs using machine learning (ML) and deep learning (DL) follows a structured process, beginning with sequence collection from in vitro and in vivo experiments. Key features such as hydrophobicity and secondary structure are extracted, and classification models categorize peptides based on their properties. ML models predict anticancer effectiveness, followed by toxicity checks and Structure-Activity Relationship (SAR) analysis to ensure safety and efficacy, with validation tests confirming their activity. This review explores how the automated design of ACPs can be enhanced by leveraging advanced ML and DL techniques. These methods, with their ability to automate feature selection and activity prediction, have significantly improved the efficiency and accuracy of peptide discovery. This structured approach holds high potential to guide researchers in the automated design of ACPs, accelerating the discovery of effective peptides while ensuring safety. Special attention is given to new approaches such as Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and Generative Adversarial Networks (GANs), which show promise in addressing key challenges like data imbalance and computational complexity. Moreover, we examine the latest published research to compare the performance of various ML models in ACP prediction. By considering these advancements and challenges, this review outlines future opportunities for improving the scalability and reliability of ACP discovery using AI-driven techniques. This structured approach underscores the transformative impact of automation in peptide design, pushing the boundaries of modern cancer therapy development.</p></div>","PeriodicalId":8449,"journal":{"name":"Artificial Intelligence Review","volume":"58 5","pages":""},"PeriodicalIF":10.7,"publicationDate":"2025-03-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10462-025-11148-3.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143554007","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}