After a review of the records, 9 categories were developed post hoc to describe trends in Korean medical device development (Figure 2). The 9 groups included digital therapeutics; virtual reality (VR) and artificial intelligence (AI); machine learning and robot; gaming and visual contents; eye-feedback and movement intervention; electroencephalography and neurofeedback; cognitive behavioral therapy (CBT) and working memory; diagnosis and rating scale; and musical, literary therapy, and mindfulness-based stress reduction (MBSR). These 9 groups were recategorized into 2 big groups: digital therapeutics and traditional therapy.

Table 1 summarizes the selected resources on the trends in Korean medical device development for ADHD and ASD.

Digital therapeutics is a broad category that refers to high-quality software, that is, digital technologies, including AI, VR, augmented reality (AR), apps, and wearable devices, that provide evidence-based therapeutic interventions to patients to prevent, manage, or treat medical disorders or diseases [15]. Digital therapeutics is one of the promising methods of intervention, treatments, and diagnosis for ADHD and ASD in South Korea. Two dissertations [13,14] and 1 research report [15] described digital therapeutics. Several specific types of digital therapeutics, such as AI, machine learning, and VR, will be further discussed in detail in other groups later.

The digital health care market is consistently growing in South Korea, especially since the COVID-19 pandemic has facilitated web-based treatments and health communication. The South Korean government has suggested diverse policies to promote the digital health care industry. In April of 2022, 10 digital therapies were approved by the Ministry of Food and Drug Safety of the Republic of Korea to conduct clinical trials [15]. However, none of them have been approved as a medical device [15]. Although the Korean digital therapeutics industry is in a beginning stage of development, the digital therapeutics market is highly promising in South Korea given the policy support and attention from the South Korean government.

For instance, digital therapeutics is applied to treatments and education for students with ADHD [13,14]. Digital therapeutics is emphasized as a new treatment approach for children and adolescents with ADHD. A substantial improvement was found in groups using digital therapeutics compared with control groups [14]. Educational materials about digital therapeutics were also highlighted for elementary school teachers educating students with ADHD. The use of Korean medical devices for ADHD and ASD reflects the attention to educating teachers about the importance and functions of digital therapeutics, as well as its direct application to those with ADHD or ASD.

VR and AI are categorized into digital therapeutics. Both technologies are promising tools that recent studies have highlighted their potential [19,20]. VR and AI were mentioned in 11 resources. Two government research and development project reports [18,26], 6 peer-reviewed articles [10-15,18-23,26], 2 dissertations [16,20,24,25], and 1 interview [17] described medical devices for ADHD and ASD using VR and AI.

Although the fundamental treatment method for ADHD is medication, behavioral problems are treated by CBT [17]. However, traditional CBT has limited accessibility in clinical settings due to additional time to visit hospitals, health care personnel, and relevant resources [17]. In this circumstance, VR technology has a big advantage in solving these limitations by enabling real-time simulations and virtual training [17]. VR-based diagnosis of ADHD is also considered to have accurate and objective results given that the model is based on VR settings, while the traditional diagnosis relies on verbal interviews [25]. Furthermore, virtual social interactions allow repetitive practice for anger recognition, anger regulation, and social problem solving [22]. This VR-based training helps control their aggressive and impulsive behaviors [22]. Some recent studies also indicated that VR-based interventions for ADHD can prevent potential crime, especially for young people with ADHD [22,23]. An AI-automated diagnosis system for diagnostic classification and automated detection based on the biosignals of ADHD was introduced for the prediction, suppression, and prevention of adolescent recidivism [23].

To enhance the efficiency of treatment and diagnosis of ADHD and ASD, VR technologies are often integrated with AI to predict, analyze, and define different types of data from patients and users [19,20]. For example, a device using a gradient-boosted decision tree algorithm was evaluated to test the accuracy of its AI-based software when health care providers diagnose ASD in children aged between 18 and 72 months [19]. The study found that an increased number of children with ASD were able to be diagnosed in a primary care setting, potentially promoting early intervention and treatment [19]. In addition, the combination of a deep learning algorithm and facial images is a novel approach in the diagnosis of ASD [16]. Given that ASD is marked by impaired neurological development, the human face provides insights into brain structure and function [16]. Consequently, facial features can serve as an important biomarker for diagnosing ASD [16]. This idea is also applied to a wearable digital intervention that promoted emotion recognition and facial engagement [24]. Researchers found that children with ASD who wore superpower glasses showed significant improvements on socialization by providing social cues and detecting facial expressions [24].

Following the trends of the combined technologies with VR and AI, research and government research and development projects have studied possible medical device models for people with ADHD or ASD [18,21,25,26]. The Korea Electronics Technology Institute conducted a project to develop VR and AR platform technology based on biosignals for the mental health of kids and silver generation [18]. In total, 7 types of VR, 6 types of AR mental health content, a mental-care cloud platform, wireless transfer technology for 4K video streaming, and systems to measure and analyze biosignals were developed throughout the project [18]. These VR and AR technologies can be broadly applied to mental health VR and AR platform services at mental-health clinics in South Korea [18]. This application also positively impacts the Korean web content industry beyond the Korean medical device industry.

Medical VR and AI technologies were interweaved with IT and life care content markets [21]. Regarding ADHD treatments, contents and systems using immersive and vivid exposure in virtual settings have been actively tried [21]. The system virtually provides the actual circumstances where people with ADHD can be trained through sensory, cognitive, and linguistic simulations [21]. With a similar purpose, the Office of Research Affairs at Yonsei University conducted a project to develop mobile VR neuropsychological batteries and an AI-based database of early diagnosis and promotion systems using digital phenotypic modeling [26]. The developed device was based on a multilayer platform integrating emotions, social ability, and neurological information [26]. Both devices target vulnerable populations with limited access to traditional treatments for ADHD.

The overall trends of Korean medical devices for ADHD and ASD concentrate on improving the current conditions of the medical device application and its use [25,26]. While most of the traditional approaches require personnel, physical resources, and travel time, VR and AI–based medical devices minimize the requirements.

Machine learning and robot-based medical devices for ADHD and ASD are also included in the category of digital therapeutics. They were found in 6 resources. Two dissertations [29,32] and 4 journal articles [27,28,30,31] addressed its trends.

The importance of early diagnosis is highlighted in many studies about ADHD and ASD [30,41]. To facilitate early diagnosis and ADHD screening, machine learning and robot-based technologies are used as a promising tool. A machine learning predictive model is one of the solutions to increase the accuracy of ADHD prediction [30]. As a longitudinal predictive model, several types of machine learning analysis were applied to predict the future and classify findings, such as supervised learning, random forest, gradient boosting, and neural network models [30]. This model identified that children who showed specific risk indicators during infancy and early childhood are likely to be diagnosed as being at risk for ADHD when entering elementary schools [30]. Similarly, machine learning and a wearable biosensor help to predict imminent aggressive behavior in inpatient young people with ASD [27]. In addition, several robot-based ADHD screening devices have been tested, such as a contactless sensing system, a deep learning–based classifier, storyboard content for children, and an automated childhood ADHD classifier [29,31,32]. The contactless sensing system, for instance, quantitatively measures the movements of children with ADHD [29]. These devices automatically detect and analyze behavioral reactions, and identify results based on collected data [29,31,32].

Furthermore, machine learning and robot-based devices are also applied to interventions. Remote robot–based interventions are effective in enhancing the level of concentration and encouraging positive learning attitudes among children with severe ADHD symptoms [28]. They recognize a robot as a peer, a good behavioral model, and a learning helper [28].

The overall observations and findings imply that robot-based models are relatively more attractive in younger ages. Machine learning systems also have a higher effectiveness and accuracy of screening.

Many types of gaming can be a part of digital therapeutics, depending on their medium. Given that a gaming approach has less rejection than others [35], its use is actively discussed in the recent medical device development for ADHD and ASD. Gaming was mentioned in 4 resources. One dissertation [35], 1 interview [33], 1 journal article [34], and 1 research report [36] found their trends in Korean medical device development for ADHD and ASD.

Gaming is applied to various fields today, not just as an entertainment tool [36]. The research found that gaming helps people relieve negative emotions and improves symptoms [36]. The development of gaming items was motivated by one of the limitations that psychiatrists’ diagnoses rely on subjective individual decisions [33]. A gaming device, AttnKare, made by Hippo T&C, is complex equipment that uses a VR test and measures eye movements and patience [33]. The AI in the device analyzes the collected information and makes individual diagnoses [33].

The cognitive rehabilitation field recently uses computer technology, focusing on basic cognitive function, memory, problem-solving ability, and perception of space and time [35]. This new digital model is personalized to different individuals [35]. Serious games in this field, defined as education-purposed games with entertaining functions [36], are a promising method that can result in easier and faster positive outcomes both in education and treatment [35]. For example, when comparing responses from 2 groups using a communication-functional board game or a traditional board game, those who used a communication-functional board game showed a better score in self-control, self-esteem, family function, and peer relationship [34].

Eye-feedback and movement technologies are often found in ADHD screening devices. Because eye movements are linked to brain areas with neuropsychological functions, such as response inhibition, selective attention, and working memory, their impairments lead to the primary traits of ADHD [41]. Eye-feedback and movement intervention also have a complex relationship with the categories above, such as VR and AI, gaming, and machine learning. The information about this category was found in 5 resources. Two dissertations [37,38] and 3 journal articles [39-41] described medical devices using eye-feedback and movement intervention.

Using a screening model for ADHD with eye-tracking features and machine learning, 33 eye-tracking features were identified to distinguish children with ADHD from developing children [41]. Eye-tracking characteristics have the potential to serve as a reliable marker for compromised neurobiological function in individuals undergoing ADHD screening [41]. The focus reaction time tests were identified as a valid tool for diagnosing children with ADHD [40]. Given that visual materials tend to be eye-catching and vision accounts for 80% of human recognition [37], visual content can also play an important role in developing interventions for ADHD using eye movements. For example, eye-feedback training improves sluggish cognitive tempo, one of the symptoms of ADHD that shows a lack of energy, slowness in behavior or thinking, and drowsiness [38]. A motion-training system with real-time visual feedback also facilitated motion control in children with ASD [39].

This category discusses electroencephalography and neurofeedback. Both concepts are relevant to digital technologies, including gaming and machine learning [49]. Electroencephalography, a recording of the brain’s electrical activity, measures brainwaves. Neurofeedback is used to modify brainwaves by providing stimulus in neurofeedback training, which is considered a promising physiological approach for the diagnosis and interventions of neurological disorders, such as ASD and ADHD [42,45,47,48,52-56,60]. This topic was mentioned in 19 resources. Two research reports [44,45], 10 journal articles [42,43,48,50,51,55,56,58-60], and 7 dissertations [46,47,49,52-54,57] discussed electroencephalography-based medical devices.

Wearable wireless systems and sensing systems are new potential solutions for diagnosing ASD and ADHD by collecting physiological indicators [42,43,59]. Electroencephalography can detect the abnormalities of the neural system related to ASD and ADHD [42,43,59]. The research found that ADHD can be diagnosed by sounds derived from brainwaves, using (1) ADHD diagnosing algorithms developed by electroencephalography brainwaves with several mathematical methods, eyes-open, and resting-state brainwaves, and (2) a sonification algorithm to convert brainwaves to musical sounds [46].

Convolutional neural network (CNN) is another emerging idea to automatically extract electroencephalography features for medical diagnosis [48,49]. CNN is a neural network modeled after the functioning of the visual cortex for processing data that contains spatial information [49]. Recent research explores deep learning–based devices using CNN to effectively classify electroencephalography signals [48]. A deep learning–based approach using functional magnetic resonance imaging (fMRI) was also another recent discussion [49]. While previous trials covered the entire brain area to identify ADHD, the recent study suggests examining specific brain portions related to the classification of ADHD using the deep learning system by demonstrating a higher level of accuracy [49].

Neurofeedback is another key topic in interventions for ADHD and ASD. Neurofeedback training is a form of self-regulation therapy for brainwaves, using the concept of operant conditioning [52]. During brainwave measurement, patients receive visual or auditory feedback on cortical activity to normalize brain function by inhibiting or reinforcing specific frequency ranges of brainwaves [52]. Neurofeedback enables them to receive real-time feedback on their brainwave states and engage in training to regulate brainwaves as desired [52]. Many studies claim that neurofeedback training positively impacts children with ADHD [47,52-54,60].

Neurofeedback training positively impacted children with ASD by improving their attention and abnormal brainwaves [52]. Students who received neurofeedback training showed increased scores in memorizing numbers and matching colors, numbers, and words [52]. Furthermore, neurofeedback training can be applied to those with ADHD [47,60]. Recent research has reported that 30% of people with ADHD with executive function deficits and inhibitory deficits cannot be treated both by medication and CBT [50,60]. Neurofeedback training is suggested as one of the promising alternative solutions of medication to improve executive functions, inhibition, and working memory [47,50,60]. Moreover, delayed language development and communication ability among children with ADHD can be improved by neurofeedback training [53,54]. In fact, parents having children with ADHD have reported positive outcomes after using neurofeedback training [50]. These trends imply that neurofeedback models can be more effective when they are integrated with different digital items, such as VR, gaming, and AI [54].

As an example of complex medical devices for interventions for ADHD, research suggested a robotic intervention education using neurofeedback [55]. In this program, students with ADHD were encouraged to craft a robot and control its movements using brainwave signals [55]. This program aimed to enhance the level of concentration as well as treatment of ADHD with a children-level storyline [56]. The satisfaction was evaluated positively, while a general operation process had a few comments on further development [56]. In addition, CNN is also used to diagnose ADHD in young children <16 years who are too immature to perform self-diagnosis or use medical equipment [51]. Gaming content is used to increase the objectivity and accuracy of ADHD diagnosis and collected electroencephalography data are classified based on the CNN model [51].

With a similar context of education using neurofeedback, the Korea Institute of Curriculum and Evaluation conducted a 2-year project to design and implement brain-based training for children with ADHD [44,45]. Neuroeducation was applied to the project to explore neuroeducational research tools, including electroencephalography, positron emission tomography, and fMRI [44]. fMRI was especially highlighted to indirectly measure brain activity status by quantifying cerebral blood volume, cerebral blood flow, and blood oxygen saturation [44]. The training program, named Korea Institute of Curriculum and Evaluation Working memory Enhancement Program, involves altering brain function through interaction with the environment, which leads to improved cognitive functions [44,45]. The Korea Institute of Curriculum and Evaluation Working memory Enhancement Program showed positive outcomes among children with ADHD in a clinical trial by enhancing cognitive abilities and demotivating behavioral problems [45].

Traditional therapy mostly does not use medical devices. The 3 categories under traditional therapy examined the trends in traditional treatments and diagnosis for ADHD and ASD.

CBT and working memory fundamentally aim to improve cognitive ability as well as attention deficits and impulsive behaviors [61,63,64]. CBT focuses on a behavioral intervention [61], while working memory refers to a cognitive function that involves temporarily holding or manipulating information for a short period [62,63]. They were found in 5 journal articles [61-65].

CBT can be a more effective intervention for adults with ADHD than children with ADHD because adults relatively have a higher cognitive ability and reflective thinking [61]. CBT demotivated people to think about ADHD and think negatively, while knowledge of ADHD was increased [61]. CBT can also be developed as a self-monitoring cognitive training program to help children with ADHD regulate and monitor their thoughts and behaviors during task execution [64]. This approach focuses on individual behavioral problems as well as individual thinking processes, which can be applied to diverse treatments and research on ADHD [65].

A working memory training program is another method to reduce impulsive behaviors and hyperactivities [63]. Given that delivery forms of information and cognitive ability are correlated with one another, previous research findings indicated that delivery forms of information influence outcomes of working memory training programs [62]. This statement implies that a better performance is presented when performing a preferred delivery form of information [62], meaning that understanding a target population’s preference for communication matters in working memory training.

While many studies discuss the recent trends in medical devices for ADHD and ASD, mostly focusing on digital technologies, traditional methodologies of diagnosis and rating scales are still discussed to update the standards and guidelines. The diagnosis and rating scales were examined in 10 resources. One news release [73], 7 journal articles [1,66,68-72], and 2 dissertations [67,74] addressed the current trends in diagnosis and rating scales of ADHD and ASD.

ADHD diagnosis should be systematically approached through diagnostic algorithms to make safe and accurate decisions [70]. Multiple factors, including age, gender, and individual perceptions of ADHD, need to be considered, and the diagnostic decision needs to be based on the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition (DSM-5) [70]. The International Classification of Diseases, 10th Revision (ICD-10) is also discussed [69,70]. Although a few differences are presented between DSM-5 and ICD-10, both models focus on attention deficits, hyperactivities, academic and social difficulties, and impulsive behaviors [70]. ADHD diagnosis usually refers to the DSM-5, while public health statistics and materials are based on ICD-10 [69]. ICD-10 has more strict standards of ADHD diagnosis than DSM-5 by recognizing all 3 categories: attention deficits, hyperactivity, and impulsion [69].

In addition to DSM-5 and ICD-10, the Children Behavior Check List (CBCL) is a self-report assessment scale developed to evaluate various emotional and behavioral problems of children and adolescents through reports from parents or close adults in their environment [67]. In South Korea, the US version of the CBCL 4-18 in 1991 was standardized and first introduced as the Korean version of CBCL, and the Korean version of the CBCL 6-18 is the recent version for parents [67,71].

ADHD screening and evaluation were performed in in-person interviews at hospitals, mental-health centers, and school counseling offices [72]. Two interview tools are used: Diagnostic Interview Schedule for Children-IV and Kiddie-Schedule for Affective Disorders and Schizophrenia-Present and Lifetime Version (K-SADS-PL-K) [72]. While Diagnostic Interview Schedule for Children-IV is a structured interview tool that can be administered by general people, K-SADS-PL-K is a semistructured tool that should be administered by trained evaluators [72]. With K-SADS-PL-K, a recent study tried the advanced test of attention, consisting of visual tests and auditory tests that present target and nontarget stimuli at regular intervals, and participants were instructed to respond only to the target stimuli [74]. However, the accuracy of distinguishing a group with ADHD from another group without ADHD was not high, which suggests limitations in using the advanced test of attention as a diagnostic tool for confirmation [74].

Given that ADHD symptoms tend to be presented at an early age, parents’ and teachers’ knowledge and perception of ADHD greatly impact their children’s diagnosis and intervention [1]. Interestingly, the ratings of parents and teachers about symptoms of children with ADHD had no significant correlations, and parents’ ratings and DISC were not matched [1]. By contrast, the rating of teachers was consistent and showed a high correlation with DISC [1]. These findings imply that DISC and the rating of teachers are more reliable and consistent than the rating of parents [1].

In case childhood ADHD may persist into adulthood, the Korean Adult ADHD Rating Scale was developed for monitoring and screening treatment of adults with ADHD [66]. Inattention was recorded as the most general symptom of ADHD in adulthood [66]. The Korean Adult ADHD Rating Scale is expected to effectively rate difficulty in emotional control and disorganization, such as inattention, hyperactivity, and impulsivity [66]. This rating scale was also suggested to extend its range of use to adolescents, embrace gender differences, and identify screening and rating scales, respectively [68]. In this light, traditional rating scales are consistently developed. For example, one of the recent rating scales is a tactile stimulation distribution device to quantify exercise and MBSR [67]. The details of MBSR and other traditional therapies are discussed in the last category.

MBSR, musical, and literary therapy described here were developed to increase the effectiveness of screening and intervention for children with ADHD. They were addressed in 5 resources. One journal article [77] and 4 dissertations [75,76,78,79] discussed how they were recently shaped.

The tactile stimulation distribution device was motivated by mindfulness, MBSR, and CBT and scientifically demonstrated a level of concentration of subjects [78]. The quantified data of stimulation were compared with the cognitive outcomes of subjects [78]. The correlative data were referred to as concentration, and the opposed data were considered a distraction [78]. This logic was also supported by left- and right-brain activities [78]. In fact, an MBSR-based program showed a significant improvement in reducing inattention, stress, and anxiety in college students with ADHD [78].

Another approach to intervention for children with ADHD is literary therapy based on social skills training [78]. The program was designed to train them to improve a social relationship between peers and adults and engage in group activities at home and school [78]. Using photo cards and photo books to inspire their imagination and creativity, the general symptoms of ADHD decreased with a significant improvement in emotional and mental stability [78].

Musical therapy is also used for screening and intervention of ADHD. Screening ADHD using musical therapy aims to strategize a plan of treatment by understanding individual conditions and the goals of treatments [79]. This screening is essential to comprehend how musical reactions can be used to improve symptoms when music attracts clients’ changes [79]. While musical therapy screening is designed for a broad understanding of individuals, interventions using musical therapy have a specific purpose to target specific symptoms. A rhythm-based musical intervention was developed to enhance timing control in children with ADHD [75]. The protocol contributed to controlling motor timing and perceived timing using a metronome, guiding a proper speed of response to suggested stimuli in the environment [75]. Carl Orff’s pedagogics, focusing on improvisatory performance with observation, imitation, exploration, and music literacy, is another type of intervention using musical therapy [76]. This program required small group activities, which encouraged social interaction with different individuals [76]. Furthermore, improvisatory work improved inattention, and imitating musical expression demotivated hyperactivities [76]. These findings indicate that musical therapy is a highly effective method both for screening and intervention in children with ADHD.