Application Strategies of Brain-computer Interface in Education from the Perspective of Innovation Diffusion Theory

Abstract

Diffusion Innovation theory aims to elaborate the law of diffusion and development of innovation things in social systems and is widely used in the study of diffusion phenomena of new technologies at home and abroad. As digital transformation reshapes global communication and learning styles, the recent advances in wearable and portable brain-computer interfaces (BCI) provide an opportunity for practical implementation in educational scenarios, have a broad application prospect. Here, we review the four ways BCI is currently applied in the field of education based on the functional classification of BCI. Next, specific factors of the innovation diffusion process of BCI in the education field are discussed within the framework of the Innovation Diffusion theory, and summaries the commercialized products of BCI in education under the four application modes, and the development route of BCI in education in the future. Based on the collected data, we believe that the current application of BCI systems in the education field is still in the initial stage of innovation diffusion, with slow diffusion. Finally, several strategies aiming to accelerate the application of BCI in the field of education are proposed and discussed, which we are hoping to have both theoretical and practical significance.

Keywords:

• Brain-computer interface (BCI)
• Education
• Innovation Diffusion Theory
• Educational Digitalization

1. Introduction

Brain-computer interface (BCI) is a technology that recognizes an individual’s thoughts by collecting and extracting brain signals, facilitating direct interaction between the brain and internal or external devices [1]. The brain contains over 100 billion neurons responsible for various cognitive functions, including reasoning, planning, and thought processing [2]. These neurons serve as the fundamental units of the brain, forming intricate neural networks through interconnectivity [3] and linking long-term and short-term memories [4]. These networks can reorganize themselves in response to environmental changes, affecting learning ability and cognitive development [5,6]. In the first two decades of the twenty first century, the fundamental framework for information technology (IT) in education was established [7]. This framework provides the foundation for integrating modern intelligent technology, facilitating the sharing of digital resources, and enhancing the quality of teaching and learning. Leveraging this foundational setup is a strategic choice to improve the quality of education [8].

At the turn of the century, scholars described the gap between neuroscience and educational applications as “a bridge too far” [9]. This was due to the lack of tools to observe brain activity in real time during and around the educational process. The advent of BCI has now provided such tools. In recent years, the performance of BCI systems has been further enhanced through the design and implementation of paradigms and algorithms [10]. The development of BCI wearability and portability has enabled the use of BCIs in non-experimental settings [11], providing an opportunity for practice and training in educational scenarios. Currently, scholars in China and abroad have conducted a series of studies on the functions of BCIs in educational scenarios. However, BCI technology in education is still in the initial stages of application. In other words, a limited number of innovators have recognized the existence of an innovation and its functions. It is insufficient to motivate the transition to persuade organizations to adopt the innovation [12]. Research on BCI applications in education remain at the stage of technological innovation and is primarily discussed subjectively by a few early innovators. There is a lack of discussion on the factors that drive the diffusion and adoption of these applications in education.

In order to address this gap, we propose an innovative approach that incorporates a diffusion perspective into the discussion. Classical theories in communication include the technical acceptance model, the theory of planned behavior, and the theory of diffusion of innovations. These theories explore different aspects of information dissemination and acceptance, but with different research foci. The Technology Acceptance Model (TAM) is more concerned with the attitudes and behaviors of individuals towards specific behaviors than with broader sociocultural changes. In contrast, the Theory of Planned Behavior (TPB) is concerned with the attitudes and subjective norms of individuals in order to predict their likelihood of adopting a given behavior. Diffusion of innovations theory is concerned with the process of diffusion, which encompasses the acceptance of innovations, the factors that influence them, and the adoption behaviors of different groups. It also examines the evolutionary process of the diffusion of new ideas, things, and technologies through the social system.

Innovation Diffusion theory, developed by American communication scholar Everett Rogers in the 1960s and subsequently refined, offers insights into the principles governing the diffusion and adoption of new ideas, technologies, and products. The theory offers a number of advantages when studying the application of new technologies. These advantages can be broadly classified into four main areas: First, the Diffusion Innovations theory employs a process perspective, which allows for the identification of the dynamic diffusion process of innovations and the different stages and barriers to technology application. Second, the theory emphasizes the social system, focusing on the roles of different groups within it. This helps to understand the degree of acceptance of new technologies by different groups of people and the reasons for it. Third, broad applicability: the Innovation Diffusion theory is not only applicable to the field of technology, but also to other fields, which makes it a universal theory for studying the application of new technologies. Fourthly, characterizing factors: the innovation diffusion theory takes into account the characterizing factors of innovations, which helps to explain why certain technologies are more readily received.

In conclusion, the Innovation Diffusion theory provides a comprehensive framework that can assist in the understanding of the diffusion and application of new technologies in society. This theory has been widely applied in studying the diffusion of innovations within social systems, such as the public acceptance of self-driving cars [13] and the digital transformation of industrial enterprises [14] It examined the diffusion of automated parcel stations [15], explored the cross-channel integration of retailers [16], investigated the influence of acceptance and adoption drivers on smart home usage [17], analyzed the future of Virtual Reality (VR) news in the age of 5 G, Wang and Qi presented a comprehensive study on the application of IT [18]. and other related influences and strategies were discussed. The theory regards innovation diffusion as a fundamental social process where in innovations and applications gradually diffuse over time among members of a social system through certain channels. Thus, it sets the stage for researching how innovations are accepted and applied [19]. Innovative behaviors are adopted at a faster rate when the recipients perceive the innovation as relatively superior, compatible, experimental, observable, and variable in thinking [20]. This theory offers unique advantages in studying the progress of new technologies application. Moreover, it is applicable to discussions regarding the diffusion of BCI. Additionally, it aids in uncovering the pivotal factors that drive and influence the diffusion of this technology within the realm of education. This research contributes to a deeper understanding and offers a theoretically and practically meaningful framework. Here, we present strategic suggestions for promoting the application of BCI in education, grounded in an analysis of literature and case studies. The analysis evaluates the current applications of this technology in the field of education from four perspectives and leverages the principles of innovation diffusion theory to outline a framework for its future development.

By setting “brain-computer interface” and “education application” as key topics, this study first employers the Web of Science Core Collection and also conducts searches on web pages for news related to BCI educational applications. Second, based on the collected data and application cases, four pathways for BCI application in education are identified. In contrast to its widespread use in the medical field, BCI applications in education primarily involve subjective discussions by a few innovators. These innovators introduce new methods and concepts from outside the system, indicating the initial, slow stages of innovation diffusion. In the framework of Innovation Diffusion theory, this article analyses specific factors in the diffusion process of BCI in education, including relative advantage, compatibility, complexity, visibility, observability, among others. Final, the article proposes and discusses several strategies to accelerate the application of BCI in the field of education. These strategies include fostering interdisciplinary talents, standardizing ethical reviews, enhancing device usability, conducting large-scale pilot tests, and promoting open data sharing. It is hoped that these strategies will have meaningful implications for both theoretical advancement and practical application.

2. Implementation and development of BCI

BCI is a new type of human-computer interaction that directly establishes a neural communication and control channel between the human brain and the external environment. This interface is multi-disciplinary and spans several fields, including neuroscience, cognitive science, computer science, control and information science and technology, and medicine [21]. The BCI system comprises four parts: brain signal acquisition, decoding, execution, and feedback. Its working principle illustrated in Figure 1. BCIs are categorized into invasive and non-invasive types based on the signal acquisition methods. Invasive BCI systems require surgical implantation, while non-invasive BCI systems collect brain signals through electroencephalography (EEG), near-infrared brain function imaging (fNIRS), magnetoencephalography (MEG), functional magnetic resonance imaging (fMRI), etc [22]. When the brain actively engages in certain cognitive activities, action consciousness, or is passively stimulated by external stimuli, its neural electrical activity is affected, resulting in corresponding changes. These active or passive neural responses create distinct spatiotemporal patterns of brain electrical signals. By acquiring and preprocessing these brain signals, these changes can be detected and converted into features [23]. Utilizing these features, the subject’s cognitive activities can be monitored and analyzed, reflecting their real-time cognitive state. Alternatively, these features can be transformed into control commands for external devices, enabling direct control of those devices [24].

Figure 1. Single flow diagram of BCI.

Since the concept of BCI was described by Dr. Grey Walter in 1964 and proposed in 1973, it has received increasing attention until John Chapin implemented a rat to locate a robot arm by pressing a lever in 1999 [25–27]. In 2000, researchers successfully implemented invasive BCI in non-human primates and rats [28]. By 2010, this technology was extended to human volunteers [29]. In the field of non-invasive BCI, significant progress has been made in various applications, including fast character input [30], self-limb control [31], two-dimensional cursor control [32], free cursor movement [33], multidimensional mechanical arm control [34], and other tasks. In the field of invasive BCI, Philip Kennedy implanted the first invasive BCI into the human body in 1998 [35]; John Donoghue launched a BCI game called “BrainGate” in 2003 [36]; the Zhejiang University team has built an invasive BCI platform from rats to monkeys to humans [37].

There are three basic paradigms of non-invasive BCI. The first is the P300 paradigm. Farwell LA and Donchin E team proposed an event-related potential (ERP) BCI system in a 1988 paper, enabling computer communication through P300 signals [38]. In 2008, Krusienski et al. enhanced the performance of the P300 speller using stepwise linear discriminant analysis (SWLDA) [39]. In 2019, a team from East China University of Science and Technology further improved P300 speller performance through multiple grayscale flashes [40]. The second paradigm is motor imagery. In 1993, Gert Pfurtscheller’s team at Graz University of Technology in Austria demonstrated that BCI systems could classify EEG patterns online and use them to control a cursor [41]. Research teams led by Gao Xiaorong at Tsinghua University, Pei Xiaomei at Xi’an Jiaotong University [42], and Chen Qiang at the University of Science and Technology of China have conducted extensive research on motor imagery BCI algorithms and experiments, achieving significant results [43]. The third paradigm is steady-state visually evoked potentials (SSVEP). In 1992, Sutter’s team in the United States proposed a BCI for character input via visual induction [44]. In 2015, a team from Tsinghua University developed a high-speed SSVEP BCI [45]. During this development, various teams have proposed hybrid or multimodal BCI systems, which have become a fourth category. For instance, Pfurtscheller’s team proposed a BCI based on both motor imagery and SSVEP [46], and a team from South China University of Technology proposed a BCI based on motor imagery and P300 [47]. The fusion of multimodal EEG information has become an inevitable trend in BCI system technology [48].

2.1. Functional classification of BCI in education

Based on the convenience and needs of usage scenarios, non-invasive BCI systems have been applied in the field of education. In recent years, both Chinese and international scholars have explored the potential applications of BCI in education. For example, George Papanastasiou highlights BCI's potential for students with low levels of disability and healthy students, focusing on four key areas: memory and attention, spatial and visuospatial skills, collaboration and communication, and creativity and emotional skills [2]. Christopher Wegemer found that BCI applications are not only related to the treatment of physical disabilities and neurological diseases but also enhance learning and cognition, and assist in learning [49]. Chinese scholars have offered different perspectives based on application scenarios. Some believe that BCI research in education spans seven areas: reading environment research, learning material presentation, interactive behavior research, education and entertainment, online learning, sports skill acquisition, and enhancing learning performance [50]. Liu Xinyu summarized BCI's educational potential in predicting learning outcomes, cultivating self-control, identifying learning styles, detecting learning status, optimizing learning ability, evaluating teaching quality, assessing online education, and assisting special education [51]. To facilitate future technical development, we propose a classification method based on BCI application functions. Through literature analysis and case studies, this article identifies the current primary focuses of BCI applications in education as follows:

2.2. Attention detection and intervention

Attention plays a crucial role in the learning process of children. Maintaining and enhancing students’ attention is a significant research area at the intersection of psychology, education, and neuroscience. Immersion theory suggests that sustaining and enhancing students’ attention can stimulate their interest in learning, boost motivation, and thereby improve learning outcomes [52]. The ARCS model posits that attention, as a visual representation of students’ learning state, is essential for achieving classroom motivation [53]. A key factor in effective learning is maintaining a high level of attention throughout the learning process, which ensures efficient learning and optimal information reception. Research by scholars such as Gevins and Askew indicates that data from BCI can extract information about the user and infer mental states, such as workload and alertness, thereby measuring cognitive processes [54–56]

In traditional classroom settings, students’ attention gradually decreases [57]. Some teams have made relevant attempts using BCI technology and achieved positive results. Jozsef Katona and Attila Kovari developed a BCI capable of measuring average attention levels, which they evaluated through student questionnaires and exam scores. They also assessed the effectiveness of project-based learning [58]. Spüler et al. developed a digital learning environment that automatically adjusts the difficulty of exercises by detecting the student’s workload, keeping the cognitive workload at an optimal level for each individual. This approach helps meet individual needs and customizes a relaxed learning environment for efficient learning [59]. Kuo et al. proposed an attention facilitation mechanism based on EEG signals. This mechanism was designed to collected students’ attention levels and academic performance during English courses, assessing their learning attitudes and perceived usefulness. It maintained students’ attention by providing assistance when attention declined, thereby improving their English listening performance [60]. Chen et al. combined attention levels and cognitive load scales to compare initial learning with review sessions, finding that students’ attention and cognitive load were higher during initial learning than during the review stage [61]. Mohammed Serrhini conducted an experimental study on college students, revealing a significant improvement in their attention levels from the first exam to the third exam. Cai Su et al. found that real-time attention feedback can enhance students’ sense of participation and self-efficacy in science research inquiries when utilizing Augmented Reality (AR) science inquiry tools. This, in turn, positively impacts students’ science inquiry processes and greatly contributes to their science learning performance [62]. Andujar et al. utilized BCI to assess and capture learners’ reading processes, emphasizing that active reading participation can optimize teaching effectiveness and enhance reading quality. In terms of practical application, a school in Jinhua, Zhejiang Province, utilized a BCI headband developed by Brainco Company to monitor students’ attention levels in the classroom and transmit this data to the teacher’s computer as shown in Figure 2. The software system calculates the students’ attention status into a score, which is then communicated to parents by the teacher [63,64].

Figure 2. Classroom scene of a school from Zhejiang province in China.

BCI is used in online classroom and distance learning scenarios, merging digital technology with neurobiological signals to simulate a more immersive and engaging learning process. Mohammed Serrhini developed an advanced web application Application Programming Interface (API), constituting a BCI system tailored for online education. When students wear a NeuroSky BCI headset during online learning sessions, the system evaluates their attention levels while interacting with internet browsers. Immediate alert notifications are generated if attention levels drop below a certain threshold, ensuring optimal attention levels are maintained throughout the learning process [57]. Lin constructed a BCI system to monitor and label learners’ attention levels while engaging with online videos on MOOC platforms, enhancing learners’ awareness of their self-learning status. The EEG system’s identification of learners’ mental states assists teachers in restructuring and producing learning materials, facilitating effective collaboration between teachers and students and providing personalized and intelligent learning support to enhance students’ academic performance [65]. Lv Jinquan applied EEG-based Virtual Reality (VR) interaction methods to educational games for children, statistically proving the positive impact of VR games on attention levels [66].

In special children’s education, there are also certain usage scenarios for attention detection and attention intervention. Children with cognitive and reading disabilities, autism, and ADHD are generally diagnosed with neurodevelopmental disorders and often face communication challenges that necessitate enrollment in special schools. Due to the special characteristics of these students, teachers need to spend more effort to teach them than ordinary teachers [67]. The specialized principles of BCI technology have become advantageous in educational settings for such students. Leveraging BCI technology, George Papanastasio conducted training and rehabilitation interventions for students with neurodevelopmental disorders, yielding significant improvements in the social interaction and communication skills of autistic students, and demonstrating particular effectiveness in addressing hyperactivity disorder and enhancing learning [2]. With the BCI system, teachers can determine from the EEG signals, enabling timely adjustments to teaching strategies [68]. This approach maximizes the agency of children in special education and simultaneously provides neural rehabilitation treatment while accessing high-quality learning resources through technology. This approach facilitates mutual reinforcement of learning and rehabilitation, guiding children in special education toward healthy and fulfilling development.

Students’ attention can also be used as an element to evaluate teachers’ teaching ability. Teachers’ abilities such as teaching methods, teaching skills, classroom management and other skills play an important role in mobilizing students’ learning perception, thinking, and memory. They are directly related to students’ attention stability and learning outcomes. Real-time monitoring of students’ classroom attention levels through BCI can provide valuable feedback for optimizing teaching design and instructional strategies.

2.3. Emotion recognition and regulation

The affective BCI provides a deep understanding of human emotion. The learning styles of students are reflected in the differences in preferences they show when performing learning tasks. These differences may stem from individuals’ emotional characteristics, which help in understanding their learning styles, interest tendencies, and the adjustment of learning tasks. This, in turn, facilitates the optimization of teaching and learning. Students exhibit diverse learning styles, which manifest in their varying approaches to learning tasks. These styles are shaped by emotional characteristics. Understanding one’s own and students’ learning styles and tendencies can assist in identifying student interests, adapting learning tasks, and optimizing teaching and learning effectiveness. Several studies proposed that the emotions experienced and aroused by students during learning affect cognitive processes. They further suggested that improving learning outcomes and experience in Intelligent Tutoring Systems (ITS) can enhance operational and work efficiency [69]. Córdova employed a neural networks approach to categorize students into four learning styles: active, reflective, theoretical, and practical. This approach enabled the identification and analysis of the differences in problem-solving strategies employed by students with varying learning styles [70]. Lin integrated interactive agent technology with course instruction in an emotional learning system designed to facilitate students’ focusing and relaxation training in an EEG visualizer developer on the basis of NeuroSky. The system monitors, records, and accumulates students’ negative emotions. Once a certain threshold is reached, the system initiates a game feedback mechanism that enhances learning emotions. resulting in high levels of satisfaction among students using the emotional learning system [71]. Bryan Hernandez-Cuevas developed a web-based environment for high school students that integrates BCI, a visual programming languages, and computer science education technology. His finding indicated that the experience of using EEG device can enhance students’ interest and self-efficacy in BCI [72]. N. Jamil demonstrated that the implementation of a BCI model using BCI to enhance students’ cognitive abilities, improve their learning and cognitive skills and lead to better academic achievements [73]. Marchesi M showed the use of BCI techniques to reduce math anxiety [74]. Furthermore, Beate Grawemeyer collected real-time emotional states, analyzed physiological and behavioral patterns between students’ learning experiences, and introduced different beneficial intervention policies to improve learning experiences [75]. To automatically measure and analyze students’ learning motivation in real-time and without interruptions, Ghergulescu employed the Emotiv EPOC EEG system to permit students to engage in an educational game to measure. The results demonstrated that students’ learning motivation to learn fluctuated over time, which could be targeted to enhance the learning experience [76]. Raffaella Folgieri developed the BrainArt platform, which allows users to freely express emotions and transform them into creative insights. It offers custom shapes, symbols, and colors, and is equipped with BCI devices and neural network algorithms to classify and analyze brain signals to "paint" thoughts [77]. A company in the BCI field offers professional applications for cognition assessment, mental health, and emotional regulation for smart education, an it’s system supports student group engagement, as shown in Figure 3 [78].

Figure 3. Application scenario of cognitive psychological group training.

2.4. Science popularization and scientific research

The continuous development of innovative technologies presents a challenging task in constructing an environment suitable for IT education. The promotion, training, and application of technology on the ground necessitate the support of related professional positions and specialized technicians. This requires foundational work, including the training and reserve of BCI talents, the creation of relevant curricula, and the construction of effective learning environments for talent development. Abdelkader Nasreddine Belkacem and Abderrahmane Lakas provide students with an educational toolkit in their curriculum, which combined the theory and practice of BCI to teach students about neural engineering courses and support them in developing their own applications in a relatively short period [79]. Matteo Chiesi present an open-source framework based on Arduino called Creamino, allowing users to connect the system to Simulink or BCI-oriented tools and set up a large number of neuroscience experiments. Furthermore, the authors argue that its low cost and compatibility with open-source BCI tools make it particularly suitable for BCI research and educational applications [80]. Tomasz M. Rutkowski constructed BCI-LAB (Student Research Laboratory) for the TARA Life Science Center at the University of Tsukuba in Japan. This system, based on non-invasive BCI and combines a variety of sensory stimulation modalities, including auditory, visual, and tactile, enables computer science students to practice and experiment with multimedia environments and neurotechnology methods to solve computational neuroscience problems [81]. In Prantosh Kr. PAUL et al. discussed the cloud-based BCI systems and concluded that cloud computing facilitates the sharing of hardware, applications, and other software packages via internet tools and wireless media for general formal education and vocational training [82].

The concept of using a BCI platform as a tool to generate music has been successfully employed in the creation of performances [83]; Romain Grandchamp introduced a pedagogical approach based on BCI to facilitate a higher entertainment value in the interaction between musicians and audiences [84]. Hu Pei-Chi et al. developed an educational model that teaches students to design and build their own BCI systems by measuring their own brain activity to generate musical compositions [85].

2.5. Movement function

The motor function of BCI aims to rebuild the user’s motor function by facilitating the exchange of information between the brain and the device [80]. It is fundamental for human nature to move and engage in physical activity. Students with mobility impairments are enrolled in regular or special schools, depending on the degree of physical specialization. The special principles of BCI can help these students rebuild their physical senses, compensate for incomplete learning experiences due to the deficiencies of the sensory functions, create an inclusive and supportive environment, and experience the process and enjoyment of learning. Heidrich R developed a game based on BCI that simulates the inability to move freely due to dyspraxia. In this game, students must use EEGs to manipulate the computer to complete game tasks. This allows students to empathize with their classmates who suffer from dyspraxia and degenerative diseases, making them more likely to help such students in real classroom and school life [86]. Kosmyna N’s application of smart home control based on the Domus smart home platform BCI, where smart home devices such as lights, TVs, coffee makers, and blinds can be adjusted in the platform, is considered beneficial for subjects with motor disabilities and communication disorders [87]. This indicates that students with motor disabilities can enhance and restore their motor functions through BCI, creating an inclusive environment that is conducive to a more enjoyable, relaxing, and efficient learning experience.

The advent of scientific breakthroughs has paved the way for the integration of BCI technology into the global education sector. As the technology advances, it is gradually being commercialized, with corresponding products emerging in four distinct functional categories. These products are detailed in Table 1, which presents a comprehensive overview of the public information and statistics available on this topic. Among the products identified, those classified under the science popularization and scientific research functional scenarios exhibited the greatest number of products, followed by those classified under the attention detection and intervention functional scenarios.

Table 1. List of current BCI products in the field of education.

Functional	Country	Company/Brand	Product range	Scenarios and roles
Attention detection and intervention	China	CUSOFT	Brain 360	Student Mental Health Assessment and Counseling Services
	Korea	OMNIFIT Brain	BRAINWAVE EEG
	Canada	InteraXon	MUSE-2
Emotion recognition and regulation	China	CUSOFT	Cognitive Training System	Student Cognitive Assessment and Enhancement
Science popularization and Scientific research	America	NeuroSky	MindWave Mobile 2: Brainwave Starter Kit	Curriculum Aids and Development Tools
	America	Emotiv	Epoc X
	China	Neuracle	EEG acquisition system and data analysis software
	China	HNNK	Brain-computer AI development platform
	America	Wearable Sensing	DSI-24
	America	OpenBCI	Instructional Bundle
	America	CGX	Quick-30
Movement function	China	HNNK	Brain-computer AI Smart Home Environment Control System	Assistive physical abilities for children with special needs
	China	HNNK	Brain-computer Smart mouse

Based on the four functional classifications of BCI applications in education, we believe that there will be these specific application scenarios in the future: Cognitive Behavioral Detection and Intervention for Cognitive Enhancement；Student Cognitive Assessment and Enhancement; Mental Health Interventions Evaluating Teaching Process；Brain Science Knowledge Popularization to Provide Strong Support for BCI Applications Creation a Typical Learning Environment for Children with Special Needs. As illustrated in Figure 4, the requisite technical completion levels vary across scenarios. The work on Identifying Students’ Motivation and Standardized Control Device Hardware Interface will transition from the Demonstration and Promotion phase to the Industrialization phase around 2030. The Demonstration and Promotion stage will gradually transition into the Industrialization stage, during which Wearable Capture Hardware Performance Enhancement, Discipline-Oriented Generic Wearable Capture Hardware Performance Enhancement, Discipline-Oriented Generic Models, and Identifying Students’ Interests in Learning will be addressed. Student Behavioral EEG, Big Data and Cloud Platform Construction, and SQL Server EEG will also approach the Demonstration and Promotion stage around 2035. Construction of the Big Data and Cloud Platform will commence around 2025 and will be completed in stages by 2040 to the commercialization phase.

Figure 4. The roadmap of BCI development in education.

3. Characteristics of innovation elements in the educational application of BCI

The diffusion of innovation is contingent upon the presence of innovative characteristics in a product or technology. In the context of innovation diffusion theory, the relative advantage, compatibility, complexity, trialability, and observability of a product or technology influence the speed of its acceptance by members of a system [88]. BCI technology in education is distinguished by the following elements of innovation:

3.1. Relative advantage

In the context of innovation, relative advantage refers to the advantage that an innovation has over existing ideas or technologies [88]. Among the four approaches of BCI application in the field of education, BCI offers absolute advantages over existing technology in the areas of attention detection and intervention, emotion recognition and regulation, and motor function, and the direct detection and intervene in concentration and emotion. This is due to the educational scenarios and the educational attributes of the human being, as well as the neuroscience itself, which is a kind of artificial intelligence system related to education and learning content, Consequently, it has gradually become an important part of the public science popularization [89], which is an indispensable data source and experimental platform.

3.2. Compatibility

The term “compatibility” refers to the degree to which an innovation is integrated with past experience, adopters’ needs, and the existing social value systems. There is a positive correlation between the compatibility of an innovation and the speed at which the innovation itself is accepted [88]. This is evidenced by the fact that the twenty first century education field has already realized the concrete applications of various technological tools [90]. Furthermore, the application of BCI in the field of education represents a significant innovation that aligns with the strategic trend of technology-enabled education and creation of innovative new forms of education driven by new scenarios in the new era [91]. This technology is compatible with industry needs and plays a crucial role in the national popularization and development of artificial intelligence science. In the four domains of BCI application in education, BCI technology is employed in the detection and intervention of attention, the recognition and regulation of emotion, the facilitation of motor function, the enhancement of motor function in the context of user-centered classroom feedback, the optimization of classroom efficiency, and in the scientific research and popularization hardware and software platforms. This enables BCI to transcend the limitations of technological breakthroughs and become more closely aligned with the users’ development and create needs. The acceleration of innovation and the application of science and technology in numerous areas has a profound impact on the advancement of emerging technologies, economic growth, and social progress.

3.3. Complexity

The degree of complexity of understanding and use by individuals in society is a measure of the complexity of innovations. Innovations that are easy to understand and use are conducive to effective initial diffusion and promotion [88]. In the current application scenario of BCI technology in field of education, wearable BCI equipment is straightforward and intuitive, which facilitates the visualization of complex tasks. BCI developers will modularize EEG signal acquisition, develop a BCI cloud computing platforms, and open data interfaces. Developers can download corresponding installation packages for secondary development according to their own R&D needs, which greatly reduces the complexity of the application.

3.4. Trialability

Trialability refers to the extent to which an innovation can be experimented with on an existing basis. It is more likely to be accepted if it supports being experimentation with [88]. In terms of the research and development of BCI technology in the field of education, the researchable and developable attributes of the technology itself provide convenience for Chinese and foreign development teams through scientific research cooperation, popularization of science demonstration and tournament support. Based on this pathway, the potential educational customers are provide with free resources and paid discount programs, which creates a convenient conditions to enhance the trialability of BCI software and hardware platforms, and to gradually accumulate potential educational customers. The conditions were conducive to the desired outcome.

3.5. Observability

Observability refers to the extent to which a particular innovation is observed by the general public [88]. BCI technology is demonstrated to the public in the field of education, often relies on its own attributes of popularization of science and openness. A representative institution at this stage is the South China University of Technology, which has developed wearable intelligent BCI devices and development platforms for developers to use. Additionally, high school students from a certain school in Guangdong developed a BCI intelligent medicine feeding system based on this platform for the caregiving scenarios for the disabled population. The system architecture is depicted in Figure 5. The practical innovation results were awarded the Innovation Potential Award of the Student Entrepreneurship Competition, which popularized cutting-edge scientific knowledge [92]. This aroused the curiosity of the public and helped to enhance the acceptance of potential users. Pazhou Laboratory was duly awarded the “Guangdong Provincial Youth Science and Technology Education Base” on the basis of its open visits and popularization of science in schools, as well as its active engagement in science activities for young people. This has led to a wide range of positive discussions, which have helped to shape the image of a strong country as a source of a wisdom. This, in turn, has inspired the scientific and technological team to continue innovating in the field of education, to give the BCI application more popularity and reputation more visibility and reputation.

Figure 5. The architecture of intelligent drug feeding system [62].

4. Problems and countermeasures

Based on Rogers’ long-term observations of social innovation in the U.S., he summarized that the speed of diffusion of innovation in the early, middle, and late stages of slow to fast, and ultimately tends to slow down, and the process of diffusion of innovation shows the shape of an S-curve [93]. Through an analysis of BCI technology applications in education and the elements of innovation diffusion, it’s evident that BCI offers richer solutions for diverse learners and scenarios compared to traditional methods. Various innovative elements play a role in promoting its diffusion. However, the diffusion of BCI system innovation in education is still in the preliminary stage, and there is no mature application model for large-scale popularization and promotion, so the diffusion is slow. How to effectively promote the application of BCI technology in the field of education as soon as possible to successfully cross the initial stage of innovation diffusion and enter the growth period of faster diffusion, this article believes that at this stage, the following four aspects can be targeted to deal with:

4.1. Cultivate compound talents, enhance relative advantages, and reduce diffusion costs

Innovation information can circulate freely in homogeneous or heterogeneous circles [94]. The choice of dissemination channels affects the individual’s decision to adopt or reject an innovation [95]. The persuasive effect of interpersonal communication effectively affects individual’s decision on a certain innovation [96], and the expansion of dissemination channels helps to reduce the cost of the innovation diffusion process. It is known from the previous section that BCI applied in the field of education have absolute intuitive advantages over existing technologies, and the opening of communication channels helps scientific researchers and industry practitioners to exchange information to achieve the integration point and gradually land. This requires, on the one hand, developers to fully understand the needs, pain points, and usage habits of educational methods and practice scenarios to adapt to the data analysis and integration needs of different education groups of students, and, on the other hand, educational practitioners to fully understand the needs of educational methods and practice to adapt to the different data analysis and integration needs of different groups of students. This requires interdisciplinary knowledge on both sides, and there is an urgent need to cultivate compound talents.

At present, the application of BCI in the field of education is still in the initial stage of exploring, and reserving and cultivating of compound talents can start from multiple aspects. The cultivation of talents cannot be separated from innovative practice, which requires technical and application parties to cooperate to create a practical platform to cultivate compound talents with both educational teaching literacy and BCI technology innovation thinking and ability. The source of the cultivation object can come from educational practitioners, strengthen the training of educational practitioners, and improve the cognitive ability of this group for new technologies. Second, it comes from the team of teachers, while improving cognitive ability, encouraging teachers to actively try out new technologies in teaching, cultivating typical cases and leading by example. Thirdly, it encourages and strengthens the cooperation between R&D teams and educational institutions, conducts interdisciplinary education practice and research, promotes multi-party participation in practice, science popularization services, etc., allows technical parties to learn and experience in depth in the education scene, and deeply explores pain points and needs in the scene, and improves the development ability of technical personnel for pain points and the innovation ability adapted to the education scene, etc.

4.2. Regulating ethical scrutiny and reduce usage risks

According to Rogers, increasing the acceptability of an innovation facilitates the acceptance of the innovation itself [88]. The ethical issues of concern to adopters have been highly controversial and are not exclusive to applications in the field of education. The main socio-ethical issues arising from its regular application in the field of education mainly include the following aspects: first, the issue of informed consent of students and parents for data collection and provision, which requires both students and parents to be explicit and agree to participate, which is a necessary condition for the applications of BCIs; the second is the issue of data security, BCIs collected biological data, including attention, emotion, intention, etc., have individual characteristics, which can be inferred from personality preferences or personal identity [97], and if there is any unknown theft, manipulation and misuse of the data, it will pose a potential risk of leakage and violation of the threat of personal privacy; The third is the issue of fairness and justice in education, the scientific research field has proven that the BCI technology is helpful in helping individuals to gain improved skills [98]. However, the current scope and target audience are limited, if some educational practitioners use this technology to help some students gain a competitive advantage in learning, it will seriously threaten the fairness of education; The fourth, the issue of autonomy. EEG signals in the transmission of brain thinking signals also accept the reverse transmission at the same time [71], the complex composition of the brain and the current state of the art make it difficult to identify the source of a decision, if it leads to serious consequences, it will also be impossible to determine responsibility, which in turn leads to a large number of legal issues; Fifth, the problem of risk-benefit ratio. Although BCI technology is developing rapidly, it is still limited in application due to its immaturity, and it is difficult to accurately evaluate the risks and benefits of application implementation, which also poses a challenge to the normalization of application.

To alleviate the ethical issues faced at this stage, scholars and organizations from domestic and foreign have put forward their views and solutions, and for the application in the field of education, measures can be taken to alleviate the problem in the following aspects: First, the establishment of relevant laws, regulations, and management systems to strengthen the supervision and guidance of the educational applications of BCI, and to safeguard the rights of informed consent of students and their parents; second, from data collection, data use, data storage, and data storage to data management, data desensitization and encryption according to specific situations, improving data security protection, and clarifying data management methods to protect the security of brain electrical signal data; Third, actively carry out technical popularization, ethical education and training, and safety awareness training to improve the comprehensive knowledge of BCI technology among the public and educators; and fourth, establish a BCI technical ethics committee to formulate safety standards and supervision and supervision mechanisms for the research and application of BCI technology, and to standardize the development research and application [99].

4.3. Wearable hardware, accurate and efficient decoding, and improved ease of use

The traditional BCI electrode caps must be glued with conductive adhesive, and the system requires the connection of complex hardware and need to wash hair, which is extremely inconvenient to use and not conducive to the diffusion of innovation. Brain electrical signals generally characterized by mixed signals, high noise, and high interference, and its decoding efficiency and accuracy have always been a problem that global scientific research teams are committed to solving. To make the normalized use of educational scenarios more convenient, the hardware device should be portable while maintaining performance, and minimize external interference and self-interference to ensure the stability and accuracy of the acquisition state. From the market product point of view, the market has a number of companies have launched BCI wearable products. Such as Brainlink headband from Shenzhen Hongzhili, Focus headband from BrainCo, Isimple headband from HNNK, etc. However, this type of hardware generally exists after the use of charging problems, comfort issues with wearing, and limited functionality due to too few channels. In addition, the collection equipment in the experimental environment and the wearable collection headband differ in the number and material of electrodes, and the number of channels limits the expansion of functions.

To resolve the current issue with wearable hardware, developers must consider the usage and management habits of the education scenario in the development of the device. This entails incorporating more human-centric designs while reducing the volume to make the product more user-friendly in the education scenario. Secondly, select suitable electrode materials and shapes to ensure the comfort of skin contact while ensuring the quality of signal acquisition. Third, algorithm performance should be enhanced and software functionality expanded. This can be achieved by enriching software content under limited channels and developing multi-channel wearable devices, which will provide more choices for education scenarios. In terms of software, for the complex situation of normal application in education scenarios, a multimodal data cloud platform can be constructed to converge multiple biological signals, which can be used as a comprehensive analysis basis for later EEG data analysis. Secondly, the software processing algorithm must be optimized and improved to enhance the efficiency and accuracy of brain information decoding.

4.4. Conduct large-scale pilot projects, open data sharing, and promote technology trial and observation

At present, the application of BCI technology in the field of education is limited to a few case studies, with most educational activities conducted within the framework of school institutions, which have a certain degree of planning and internalization. This results in a lack of opportunities for the public to observe this innovative achievement and to participate in trials, which restricts the innovation diffusion of technology.

To address this issue, measures can be implemented in the following areas to enhance the situation: First, conduct pilot projects on a small scale through multiple channels. Establish specific application scenarios for BCI and evaluate data collected from pilot projects to assess effectiveness, feasibility, and user experience of BCI technology. Second, undertake extensive collaboration with universities, middle schools, and primary schools to conduct educational research. Explore application effects of BCI in students of different disciplines and age groups. Evaluate learning achievements, attention, emotional states, etc. Third, establish a data-sharing platform for BCI, enabling researchers and education practitioners to share data, algorithms, and experiences. This will accelerate technology development and application. Fourth, organize BCI special events and science popularization activities. Encourage student participation in BCI experiments and applications. Collect feedback from students to gain insight into their feelings and experiences during the usage process. It is imperative to pursue continuous improvements in the accuracy and reliability of BCI, the formulation of technical standards, and the reduction of technology development and application costs. Collaborate with experts in education, psychology, neuroscience, computer science, and related fields to study the optimal timing for integrating BCI technology in education. This will promote knowledge exchange and innovation.

5. Conclusion

The advancement of artificial intelligence offers numerous possibilities for improving education quality and efficiency [100]. BCI is a novel and effective tool for human-computer interaction, enhancing information transfer and communication in various educational scenarios. This technology can significantly advance the digitization and modernization of education, demonstrating great potential in the field. Recent advances in wearable and portable BCI offer realistic opportunities for educational modernization and digitization, gaining global attention. However, BCI in education is still in its initial stage, facing challenges related to innovation diffusion and development. Targeted strategies are needed in areas such as talent training, ethical review, integration and adaptation, and open pilots. Nevertheless, significant distance remains before this technology can be widely implemented and contribute to educational reform. It is crucial to facilitate continuous advancement in the technological innovation of BCI to promote the reform and innovation of education teaching modes. Furthermore, the development of more complex talents and teams is essential for sustainable exploration of these new avenues. This theoretical-driven review clarifies the current status of BCI applications in education. While this is only a starting point for developing BCI in education, it provides a framework to encourage multidisciplinary involvement in this promising field.

Disclosure statement

No potential competing interest was reported by the authors.

Additional information

Funding

This work was supported in part by STI 2030—Major Projects 2022ZD0208900; in part by the Key Research and Development Program of Guangdong Province, China, under Grant 2018B030339001; in part by the Key Realm Research and Development Program of Guangzhou, China, under Grant 202007030007.