Background: Adaptive systems serve to personalize interventions or training based on the user’s needs and performance. The adaptation techniques rely on an underlying engine responsible for processing incoming data and generating tailored responses. Adaptive virtual reality (VR) systems have proven to be efficient in data monitoring and manipulation, as well as in their ability to transfer learning outcomes to the real world. In recent years, there has been significant interest in applying these systems to improve deficits associated with autism spectrum disorder (ASD). This is driven by the heterogeneity of symptoms among the population affected, highlighting the need for early customized interventions that target each individual’s specific symptom configuration.
Objective: Recognizing these technology-driven therapeutic tools as efficient solutions, this systematic review aims to explore the application of adaptive VR systems in interventions for young individuals with ASD.
Methods: An extensive search was conducted across 3 different databases—PubMed Central, Scopus, and Web of Science—to identify relevant studies from approximately the past decade. Each author independently screened the included studies to assess the risk of bias. Studies satisfying the following inclusion criteria were selected: (1) the experimental tasks were delivered via a VR system, (2) system adaptation was automated, (3) the VR system was designed for intervention or training of ASD symptoms, (4) participants’ ages ranged from 6 to 19 years, (5) the sample included at least 1 group with ASD, and (6) the adaptation strategy was thoroughly explained. Relevant information extracted from the studies included the sample size and mean age, the study’s objectives, the skill trained, the implemented device, the adaptive strategy used, the engine techniques, and the signal used to adapt the systems.
Results: Overall, a total of 10 articles were included, involving 129 participants, 76% of whom had ASD. The studies included level switching (7/10, 70%), adaptive feedback strategies (9/10, 90%), and weighing the choice between a machine learning (ML) adaptive engine (3/10, 30%) and a non-ML adaptive engine (8/10, 80%). Adaptation signals ranged from explicit behavioral indicators (6/10, 60%), such as task performance, to implicit biosignals, such as motor movements, eye gaze, speech, and peripheral physiological responses (7/10, 70%).
Conclusions: The findings reveal promising trends in the field, suggesting that automated VR systems leveraging real-time progression level switching and verbal feedback driven by non-ML techniques using explicit or, better yet, implicit signal processing have the potential to enhance interventions for young individuals with ASD. The limitations discussed mainly stem from the fact that no technological or automated tools were used to handle data, potentially introducing bias due to human error.
J Med Internet Res 2024;26:e57093
adaptive system (3); virtual reality (468); autism spectrum disorder (57); intervention (602); training (184); children (419); machine learning (1344); biosignal (1)
Recent research on assessment, training, and intervention applied to technologies has focused on creating complex adaptive systems [1-5]. According to Almirall et al [6], an adaptive intervention is characterized by a set of clinical decision rules that offer guidance on when and how to adjust the dosage and nature of the treatment, considering specific measures. In this way, adaptivity can be defined as the system’s capability to alter its actions in response to the preferences and needs of the user [4,7]. Moreover, adaptive systems have the potential to enhance the individualized training experience and prevent issues such as overtraining, undertraining, cognitive overload, frustration, and boredom [5]. This differs from the nonadaptive approach, where the same settings are used throughout training, or adjustment is based on settings that are unrelated to the participant’s performance [8].
Technologies, such as robots, mobile phones, and screen-based systems, provide controlled and engaging environments that facilitate adaptive training and support the development of multiple interaction abilities in a secure and predictable manner [9]. The conventional computer is the predominant hardware choice for adaptive and personalized systems [3,4] because adaptive systems have been built upon existing development tools and infrastructure designed for traditional computers and devices. This approach aims to streamline the development process, ultimately reducing the human effort and time required for implementation. In contrast with traditional computer-based therapies, virtual reality (VR) excels in promoting ecological validity by delivering immersive, lifelike experiences, thus creating a strong sense of presence and facilitating the transfer of learning outcomes to real-world situations [10-13]. VR is intended as a computer-generated simulation of an environment that allows users to interact with and experience an artificial world as if it were real. VR aims to create a sense of presence, where users feel as though they are physically present within the virtual environment, enabling them to explore the simulated space, manipulate objects, and engage with the surroundings in a lifelike manner [12]. Indeed, VR technology can achieve varying degrees of immersion, categorized into 3 levels: nonimmersive, semi-immersive, and fully immersive (depending on the capabilities of the device being used). VR offers various avenues for adaptation, including adjusting the complexity of content, tailoring evaluations, and modifying autonomous virtual agents [14]. Another form of adaptation could involve integrating system input, such as using voice commands and haptic feedback. Moreover, VR not only enables the recording of real-time information but also facilitates the integration of data collected from various devices, each dedicated to monitoring distinct psychophysiological activities [15]. Several studies have yielded significant training findings on implementing adaptive VR interfaces. Among them are interventions related to mental health or neuropsychiatric conditions, such as emotional and affective training [16], treatment of phobias [17], management of pathological stress [18], and therapy for posttraumatic stress disorder [19].
In recent decades, research has focused on using VR in the assessment, treatment, and training of neurodevelopmental disorders such as autism spectrum disorder (ASD) [13,20-25]. Studies increasingly focus on this disease due to the escalating worldwide incidence and the high demand for early interventions [26]. Moreover, extensive research has demonstrated that training with VR can lead to notable enhancements in various domains among the young population with ASD [14,21,23-25]. ASD is a neurodevelopmental condition characterized by impairments in social communication and the presence of restrictive and repetitive behaviors [27]. One distinguishing aspect of ASD is its spectrum nature, indicating significant variability in symptom severity among the individuals affected. This heterogeneity results in clinical phenotypes that differ substantially from one person to another while sharing common underlying features. The heterogeneity in symptom severity observed in ASD requires the implementation of personalized treatment approaches that target each individual’s specific symptom configuration. By designing early interventions tailored to each child’s characteristics, it becomes possible to provide targeted training and improvements in deficits commonly associated with ASD [14,28]. In this connection, adaptive VR technologies seem to have the potential to pave the path for a new generation of highly efficient technology-driven therapeutic tools for the young population with ASD.
Following this approach, a variety of ASD adaptive technologies have been proposed. Among them, Bian et al [29] presented a VR training system for improving driving skills that autonomously adapted its difficulty levels according to participants’ engagement and performance metrics. Another adaptive system was applied to a VR job interview training platform, which dynamically adjusted the conversation according to users’ responses and stress levels [30]. Research has demonstrated that using such a solution for rehabilitation yields favorable outcomes in enhancing certain abilities of children with ASD; specifically, the VR system was able to tailor the intervention in accordance with the user’s actions [31].
Given the evidence demonstrating the effectiveness of using adaptive VR systems to address deficits associated with ASD and the showcased advantages of early intervention, there arises a need to examine the technical aspects of system adaptation and the complexities of handling user data. To do this, it is crucial to dissect the components of an adaptive system generally used to train or treat ASD symptoms. The following taxonomy will facilitate the reading of this work.
A system can be adapted through different strategies, such as level-switching techniques or feedback. Level switching follows a logic of level difficulty, and the choice of switching can be based on progression or regression techniques. The progression technique can be considered the core training principle because it is necessary to increase the difficulty level to continue training certain skills, improve them, and prevent the occurrence of learning effects. Therefore, in this scenario, training is adapted through a gradual increase in difficulty level based on the individual’s abilities. Unlike progression, the regression technique allows for a finer, more flexible, and more nuanced approach, adapting the difficulty level according to the arising needs. Rooted in traditional therapy practices, this strategy allows for dynamic adjustments that increase or decrease the difficulty level based on the observed performance of the individual [8]. This would ensure that the individual is able to perform satisfactorily at an easier level before increasing the difficulty level, allowing more time to learn the content [8].