Stroke is a common cerebrovascular disease characterized by high disability and mortality rates (Soekadar et al., 2015; Cao et al., 2022b; Zhan et al., 2022; Lin et al., 2023). With the aging of the population and changes in lifestyle, the incidence of stroke has been increasing year by year, making it one of the major challenges to global health (Chen X. et al., 2022; Zhang R. et al., 2023). Stroke patients often require long-term treatment and rehabilitation training (Cao et al., 2022a; Uivarosan et al., 2022; Liu X. et al., 2023). Therefore, improving the treatment effectiveness and rehabilitation quality has become an urgent problem in the field of stroke treatment. With advances in computer technology and neuroscience, brain-computer interface (BCI) technology has been receiving more and more attention and research in the treatment and rehabilitation of stroke patients (Mane et al., 2020; Liu et al., 2022). BCI technology is an emerging technology that connects the human brain with external devices, using the neural activity of the human brain to control external devices for interaction and control, thereby giving patients a sense of autonomous control and accelerating the rehabilitation process (Gutierrez-Martinez et al., 2021; Gao et al., 2022). Supported by BCI technology, nerve regeneration and functional recovery can be promoted, improving the quality of life and alleviating the impact caused by the disease (Sinha et al., 2021; Song et al., 2022). Hence, BCI technology has significant value and potential in improving the quality of life for stroke patients and has become one of the research hotspots in related academic fields both domestically and internationally (Hu et al., 2021; Li H. et al., 2023). In recent years, the application of BCI technology in the stroke field has gained popularity, and consequently, the number of academic literatures on this subject has been continuously increasing. Therefore, in order to systematically, objectively, and comprehensively understand the current development, research focal points, and trends of this field, the method of bibliometrics has been introduced into the research of the field, helping us to quickly analyze the quantitative characteristics and laws of these literature, in order to provide direction and reference for future research. Bibliometrics is a discipline that employs statistical and quantitative methods to study, evaluate, and analyze scientific literature (Chen Y. et al., 2022). It has been widely applied in various research fields (Luo and Lin, 2021; Zhang L. et al., 2022, 2023; Zhang and Lin, 2023).

This study systematically and comprehensively sorted and analyzed relevant literature on the use of BCI in stroke research fields using bibliometrics and visualization technologies. Firstly, this paper statistically analyzes the relevant literature on multiple aspects including total amount, annual publication, citation frequency, research countries, institutions, authors, and journals, in order to achieve a comprehensive description of the research situation in this field. Secondly, based on the scientometrics method, this paper conducts an in-depth analysis of the research hotspots, topics, and trends of the literature. On this basis, the development prospects and future research directions of BCI technology in stroke treatment are discussed. Finally, visualization techniques are adopted to present the results of literature analysis through forms such as tables and knowledge maps, which improves readability and intuitiveness, and better conveys relevant research information. Through these analyses and presentations, this research aims to unveil the current research status, prominent areas of investigation, and emerging trends of BCI technology in stroke treatment and rehabilitation, in order to furnish unique insights and directional guidance for further research and practice in this field.

This study employs bibliometric methods to comprehensively and systematically examine the relevant literature on BCI technology in stroke research over the past decade. Through visualizing knowledge maps, the study presents an intuitive representation of the knowledge structure and developmental trajectory in this field from multiple perspectives. The findings indicate that BCI technology has received extensive attention and attracted increasing numbers of researchers in the field of stroke research, resulting in a prolific output of research achievements and promoting the vigorous development of this field.

As one of the leading research nations, countries such as the United States, Germany, the United Kingdom, and Spain have a relatively early average publication year for academic publications. During the early stages of research in this field, they made outstanding contributions, laying a solid foundation of knowledge for subsequent studies. In comparison, China, as the top-producing country, had relatively fewer relevant research achievements prior to 2018 and was in its initial stages of development. However, in recent years, due to strong support from national policies and funding, China has embarked on the fast lane of development in this area, producing a large number of academic research achievements, and has leapt to become the top-performing country in terms of both total publications and annual publications. Nevertheless, the ACPP of these research results ranks lowest among the top 10 countries/regions in terms of publication output, indicating the overall low research quality of Chinese publications and a failure to achieve widespread dissemination in this field. Hence, the insight gained is that striving for academic prestige and influence in a research field should not only prioritize the quantity of research achievements, but also pay more attention to their quality. Therefore, in the future development of this field, China should not only guarantee the quantity of research achievements, but also strive to improve the level of scientific research and increase the output of high-quality academic achievements in order to enhance its international influence. Germany has also made outstanding contributions to this field, with the highest TLS, indicating that it has established extensive cooperation relationships with other countries/regions, particularly the United States, Spain, the United Kingdom, Italy, Switzerland, and Denmark. Switzerland, while ranking only 10th in terms of publication output, has the highest ACPP, indicating a high research level, with academic achievements enjoying widespread recognition and citation from many scholars, thereby demonstrating a high academic influence. A highly cited review study on BCI technology and its applications in severe stroke and spinal cord injury motor rehabilitation, co-authored with other countries, was published by Switzerland (Chaudhary et al., 2016). Furthermore, it is worth noting that 80% of the top 10 countries/regions with the highest publication output are located in North America and Europe, indicating that they are the main strongholds of this research field.

The University of Tubingen in Germany has emerged as a leading institution in the field, with a notable abundance of publications. These publications have contributed significantly to early-stage research in the field, with an average time of publication being relatively early. One of their highly cited papers evaluated the efficacy of BCI training in the rehabilitation of chronic stroke, revealing that BCI training can significantly improve upper limb Fugl-Meyer assessment motor scores in 32 individuals suffering from chronic stroke who were randomly assigned to groups. Notably, this improvement in motor scores was found to be correlated with changes in laterality index of functional magnetic resonance imaging and electromyographic activity of the paretic hand, thus indicating the potential of BCI training to induce motor function improvement in chronic stroke patients without residual finger movements (Ramos-Murguialday et al., 2013). Aalborg University in Denmark not only ranks high in terms of publication output but also boasts the highest TLS, indicating its highly frequent collaborations with other institutions, particularly with the New Zealand Coll Chiropract, University of Waterloo, Imperial College London, University of Gottingen, Auckland University of Technology, and Shanghai Jiao Tong University. Among their highly cited works is a study that evaluated the effects and potential mechanisms of BCI training based on Hebbian associative principles in 22 individuals with chronic stroke using a double-blind sham-controlled design. Results revealed that precise coupling between brain commands and afferent signals is essential for the reported behavioral, clinical, and neurophysiological changes, which may become a driving principle for future BCI rehabilitation design (Mrachacz-Kersting et al., 2016). Although Tecnalia in Spain has relatively fewer publications, it ranks first in the ACPP ranking, indicating its high-quality academic achievements that have gained widespread recognition and citations from scholars, thus possessing a high level of academic influence in the field. The institution recently conducted a study examining neural plasticity in the motor network of severely impaired chronic stroke patients following a therapy based on electroencephalogram (EEG)-BCI. The results demonstrated that BCI contributes to the enhancement of the ipsilesional brain activity and proprioceptive function of the affected hand, thereby leading to a reorganization of both contralesional and ipsilesional somatosensory and motor assemblies, as well as afferent and efferent connection-related motor circuits. This reorganization facilitates the partial restoration of the original neurophysiology of the motor system, even in cases of severe chronic stroke (Caria et al., 2020). Despite ranking third in terms of research paper output, Fudan University in China shows a relatively low ACPP, indicating the need for further improvement in its research level. To address this, it is crucial to enhance collaboration and communication with domestic and international research institutions. Additionally, attention should be paid to the fact that Fudan University’s publications have a relatively late average publication year, primarily focusing on recent research areas.

Birbaumer N, from the University of Tubingen in Germany, is not only the most prolific author in the field but also one of the top 10 most cited authors. Additionally, his TC, ACPP, and H-index are also among the highest, indicating his significant academic influence and important contributions to the field. One of his highly cited studies conducted a meta-analysis and assessed the effectiveness of BCI in post-stroke motor rehabilitation, revealing that BCI induces functional and structural neural plasticity at a subclinical level (Cervera et al., 2018). Farina D, from Imperial College London in the UK, ranks first in TLS, which indicates close collaboration and communication with other scholars, particularly Jiang N, Dremstrup K, Mrachacz-Kersting N, Niazi IK, and Jochumsen M. Pfurtscheller G, from Graz University of Technology in Austria, has the highest citation frequency of any author and his academic achievements have been widely cited and disseminated. One of his highly cited studies reviewed the current state of BCI technology and the challenges to overcome for its future development (Wolpaw et al., 2002). It is worth noting that high productivity and high citation authors mostly come from North America and Europe, especially Germany and the United States. In contrast, there has not been a prominent research team with significant academic influence in the field among Chinese researchers, and there is room for improvement in scientific research strength and level, which requires continuous efforts to achieve better breakthroughs.

Frontiers in Neuroscience is one of the top 10 journals in the field in terms of the number of publications and citation frequency. One of its highly cited articles reported a clinical trial that utilized BCI training on 74 stroke patients suffering from severe upper limb paralysis, which demonstrated that incorporating BCI control into exoskeleton-assisted physical therapy can improve motor function in stroke patients with varying degrees of severity, affected regions, and duration since onset (Frolov et al., 2017). Journal of Neural Engineering is the top-ranked journal in terms of citation frequency and TLS, and is among the top three journals with respect to publications. Furthermore, it has established strong co-citation relationships with other journals such as Neuroimage, Clinical Neurophysiology, Frontiers in Neuroscience, Nature, Lancet, Science, Plos One, IEEE Transactions on Neural Systems and Rehabilitation Engineering, Stroke, Journal of NeuroEngineering and Rehabilitation, Frontiers in Human Neuroscience, and Journal of Neuroscience. One of its high-profile papers evaluated the advantages and limitations of EEG-based BCI paradigms from various perspectives and analyzed potential issues of EEG-based BCI systems, proposing possible solutions (Abiri et al., 2019). Furthermore, Neuroimage is the Q1 journal with the highest ACPP and h-index, while Stroke is the high-cited Q1 journal with the highest IF. This indicates that these two journals have high recognition and authority, and their published articles are of great academic reference value. When publishing research in this field, submitting to these highly productive journals can be a priority, and when searching for relevant literature, these highly cited journals’ collected papers can be the first choice.

Highly cited literature typically reflects high-quality research and academic impact while also providing insights into the main research focus in a given field (Wu et al., 2021; Zhu et al., 2021; Li R. et al., 2023). Therefore, analyzing highly cited literature can offer preliminary understanding of the research trends and direction in that field. A highly cited study by Pfurtscheller and Lopes Da Silva (1999) introduced the basic principles of event-related synchronization and desynchronization in EEG and magnetoencephalogram (MEG). The study found that the neural structure of the brain exhibits different responses to event-related potential (ERP) and event-related desynchronization (ERD) or synchronization (ERS), with ERD/ERS being highly specific frequency band. Daly and Wolpaw (2008) conducted a study on the application of BCI for improving motor control in patients after stroke. The technology utilizes EEG brain signals to indicate the current state of brain activity to patients and guides them to modulate abnormal brain activity, thereby facilitating the development of activity-dependent brain plasticity. Pichiorri et al. (2015) conducted a randomized controlled study on 28 patients with severe functional impairment following subacute stroke to evaluate the efficacy of motor imagery (MI) monitored by BCI on post-stroke motor recovery. The results demonstrated that the BCI group achieved better functional recovery outcomes compared to the control group, including a significantly increased probability of clinical-relevant Fugl-Meyer score improvement. This indicates that incorporating BCI technology can assist in MI practice, demonstrating the potential of MI in rehabilitation, and improving the severe motor impairment in subacute stroke patients. Buch et al. (2008) conducted a study on a BCI system based on MEG with 8 chronic stroke patients suffering from hand paralysis. The results indicated that these patients were able to achieve volitional control of the neuro-magnetic activity features recorded from the central region of the scalp after BCI training, and control gripping movements using a mechanical hand orthosis. Ang et al. (2015) investigated the effectiveness of an EEG-based MI BCI system combined with MIT-Manus shoulder-elbow robotic feedback (BCI-Manus) in chronic stroke patients with upper limb hemiparesis. The results demonstrated that the BCI-Manus therapy is effective and safe for the rehabilitation of severe upper limb hemiparesis in post-stroke patients, and the revised brain symmetry index (rBSI) could serve as an important prognostic indicator for BCI-based stroke rehabilitation. The study by Biasiucci et al. (2018) found that the combination of BCI with functional electrical stimulation (FES) is more effective in promoting motor recovery of chronic stroke survivors than sham FES. This recovery is significant, clinically relevant, and durable, and is associated with quantitative features of functional neuroplasticity. Prasad et al. (2010) investigated the role of a BCI system in providing computer game-based neurofeedback during the MI phase of a protocol to 5 chronic stroke hemiparetic patients. The results showed that BCI-supported MI is a feasible intervention that can be combined with physical practice and MI practice of rehabilitation tasks as part of post-stroke rehabilitation programs. Daly et al. (2009) demonstrated the feasibility of combining BCI with FES training for post-stroke motor learning, achieving highly accurate control of brain signals that were reliably used to trigger FES devices for isolated index finger extension. Analyzing burst literature, especially those whose burst has continued until recently, can explore research hotspots and future development trends in the field to some extent. It is noteworthy that six of the papers had burst times extending to and after 2022, indicating that the research findings of these papers may be more closely related to the latest scientific frontiers and have important reference value for understanding the future development trends in the field. Bhagat et al. (2016) demonstrated the feasibility of detecting motor intentions from the brain activity of chronic stroke patients using an asynchronous EEG-based BCI technology. By measuring the movement-related cortical potentials with an optimized EEG electrode placement, motor intentions can be accurately inferred. Successful intention detection can trigger upper-limb exoskeleton movements and guide the movements with real-time sensory feedback, encouraging active participation of the patients. Donati et al. (2016) implemented a multi-stage BCI-based gait neurorehabilitation protocol to train eight chronic spinal cord injury (SCI) paraplegic patients with the aim of restoring their motor function over a period of one year. The results showed that all patients exhibited neurological improvements in various somatosensory domains such as pain localization, fine/rough touch, and proprioception across multiple dermatomes. Utilizing electrocardiogram measurements, patients also regenerated autonomous motor control in crucial muscles below the level of SCI, leading to significant improvements in their walking index. Bouton et al. (2016) successfully restored motor function in a paralyzed human by using intracortically recorded signals to control muscle activation for the first time. Lebedev and Nicolelis (2017) conducted a comprehensive review of the research progress in BCI. The classic objectives of BCI are to reveal and exploit the operating principles and plasticity characteristics of distributed and dynamic brain circuits, and to restore the mobility and sensation of severely disabled patients through the creation of novel therapies. In recent years, BCI research has also introduced new neurorehabilitation strategies, providing new insights for clinical applications. Vansteensel et al. (2016) described a communication method for patients with advanced amyotrophic lateral sclerosis, utilizing a fully implanted BCI consisting of dura-subpial electrodes placed on the motor cortex and an emitter placed subcutaneously on the left side of the chest. After 28 weeks of electrode placement, patients were able to accurately and independently control a computer typing program at a rate of approximately two letters per minute by attempting to move the hand contralateral to the implanted electrodes. The analysis of the highly cited and burst literature presented above can aid in acquiring a foundational knowledge of the field, identifying research hotspots, and predicting future developmental trends. These high-quality publications possess significant academic impact and have played a crucial role in driving the progress of the field. Through further examination of these publications, we can gain a better understanding of the current state and future directions of the field, thereby enhancing our research capabilities and the quality of our research outcomes.

The co-occurrence analysis of keywords demonstrates that stroke, BCI, rehabilitation, MI, recovery, EEG, neurorehabilitation, and classification are the major research topics in this field. By clustering these keywords and displaying them in a timeline graph, it can be observed that the clustering labels of #2 MI, #4 BCI, and #6 excitability have continued to evolve to date, which to some extent reflects the current research hotspots. Burst keywords refer to the research hotspots that have been frequently studied during a certain period of time. In recent years, burst keywords in this field include neurorehabilitation, BCI, task analysis, stroke patient, FES, feature extraction, motor impairment, and induced movement therapy. These research hotspots not only reflect the current development trends but also provide valuable insights for future research. In the field of neurorehabilitation, research on BCI, stroke patient rehabilitation technology, and motor function recovery has become major research hotspots. In the future, further in-depth research could explore issues such as task analysis, FES, and feature extraction, aiming to improve the effectiveness of rehabilitation treatments and provide beneficial support for stroke rehabilitation practice.

This study employs bibliometric analysis and visualization techniques to comprehensively and systematically analyze the research on BCI technology in the field of stroke for the first time. From multiple perspectives, including collaboration networks, co-citation networks, and co-occurrence networks, the knowledge framework and development trajectory of this research field are presented. Moreover, the study explores research hotspots and future development trends, providing valuable references and directions for further research. However, there are certain limitations to this study. Firstly, the literature data was solely obtained from the WoSCC database, potentially leading to the omission of relevant literature from other sources. Nonetheless, it is important to highlight that the WoSCC database is widely recognized as the most commonly employed database for bibliometric research (Zhong and Lin, 2022; Dong et al., 2023; He et al., 2023; Jiang et al., 2023; Li M. et al., 2023; Liu T. et al., 2023). Additionally, some recently published high-quality literature may be underestimated. Finally, the incomplete dataset of this year has unavoidably restricted its inclusion in this study.

This study encompasses a total of 693 relevant literature sources, published by 2,556 scholars from 975 institutions across 53 countries/regions and collected by 185 journals. Notably, over the past decade, both the annual publications and citations related to BCI technology in stroke research have demonstrated an overall upward trend, indicating rapid development and significant attention in this field. China and the United States stand out as highly productive countries, while the University of Tubingen has made the greatest contributions. Birbaumer N and Pfurtscheller G, respectively, rank as the most prolific author and highest cited author, both of whom have earned high academic recognition and impact through their outstanding work. Regarding journals, Frontiers in Neuroscience published the most literature, while Journal of Neural Engineering had the highest citation frequency. The research hotspots in this field encompass crucial keywords such as stroke, BCI, rehabilitation, MI, motor recovery, EEG, neurorehabilitation, neural plasticity, task analysis, FES, motor impairment, feature extraction, and induced movement therapy. These hotspots keywords to some extent reflect the development trends and frontier research directions in this field. However, BCI technology still faces numerous challenges in stroke research, including electrode stability, complexity of signal decoding, long-term stability and reliability, adaptability to individual differences, and limitations in clinical applications. Therefore, future research should focus on improving the precision and stability of the technology, enhancing user-friendliness, increasing adaptability to individual differences, and strengthening measures for safety and risk management. Overall, this study, presented in the form of a knowledge graph, comprehensively and systematically showcases the widespread and in-depth literature resources related to BCI technology in the stroke field. This assists scholars in conveniently gaining a more profound understanding of the development and prospects of this field, thereby promoting further research efforts.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

FL, DZ, and ZH designed the study and performed software analysis. FL, JC, KT, XL, and ZH contributed to data collection and verification. FL and ZH drafted the manuscript. DZ, JC, KT, and XL revised and approved the final version of the manuscript. All authors read and approved the submitted version.

The study was supported by the Innovation Project of Guangxi Graduate Education (No. YCBXJ2022009), Guilin Science Research and Technology Development Program (No. 2020011208-5), and Self-Funded Research Project of the Health Commission of Guangxi Zhuang Autonomous Region (No. GZZC2019175).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.