The intricate relationship between motor functioning and neurodevelopmental disorders, particularly Autism Spectrum Disorder (ASD) and Developmental Coordination Disorder (DCD), has been increasingly acknowledged in recent years. This growing recognition is evident in the extensive research and literature that delve into the subtle yet significant distinctions in motor skills, kinematics, and neural mechanisms underlying these conditions.
Motor Functioning in Autism Spectrum Disorder and Developmental Coordination Disorder
Research indicates that over 80% of individuals with Autism Spectrum Disorder (ASD) exhibit notable differences in motor functioning (Bhat, 2020; Licari et al., 2020; Miller et al., 2021; Zampella et al., 2021). These motor differences are not currently included in the diagnostic criteria for ASD but are significant enough to warrant attention.
In comparison, individuals with Developmental Coordination Disorder (DCD) experience impairments in various motor skills, including fine and gross motor skills, dexterity, limb speed, and visual-motor integration (Blank et al., 2019). Unlike ASD, DCD does not include social functioning impairments as part of its diagnostic criteria, although secondary social differences may arise due to reduced opportunities in activities like sports (Cermak & May-Benson, 2020). Both ASD and DCD are characterized by disruptions in basic, postural, and purposeful motor control (Dewey et al., 2007; Eggleston et al., 2017; Lim et al., 2017; Miller et al., 2019; Mostofsky et al., 2006; Paquet et al., 2019; Roley et al., 2014).
Assessment and Identification of Motor Skills
The assessment of motor skills is crucial for identifying individuals at high risk for motor delays. However, capturing subtle coordination and timing aspects remains challenging (Campbell & Hedeker, 2001). Traditional assessment methods, though detailed, are prone to human error and time-consuming (Harris et al., 2015). Technologically advanced methods like optical motion capture provide more nuanced insights but are costly and technically demanding.
To overcome these limitations, there has been a growing interest in using motor game play coupled with machine learning for diagnosis and understanding of motor dysfunctions. Previous studies have successfully utilized machine learning with data from smart tablet motor games or wearable devices to distinguish ASD from typically developing (TD) children (Anzulewicz et al., 2016; Tunçgenç et al., 2021). However, distinguishing between similar disorders like ASD and DCD remains a challenge.
Distinguishing ASD and DCD through Kinematics and Neural Imaging
In ASD, more than 80% of individuals show motor coordination deficits, impacting movement, gait, posture, and other motor control aspects (Cavallo et al., 2021; Chua et al., 2022; Eggleston et al., 2017; Miller et al., 2019; Trevarthen & Delafield-Butt, 2013). These deficits are linked to disrupted sensorimotor integration and impaired predictive mechanisms (Chua et al., 2022; David et al., 2012; Gowen & Hamilton, 2013; Sinha et al., 2014; Trevarthen & Delafield-Butt, 2013).
Comparative studies on ASD and DCD in motor tasks yield mixed results. Some studies find no significant differences between the groups, while others show either poorer skills in ASD or DCD, depending on the specific motor assessment (Kilroy et al., 2022a, 2022b; Dewey et al., 2007; Wisdom et al., 2007; Green et al., 2002). Interestingly, ASD and DCD seem to differ significantly in the ability to imitate meaningful gestures, with ASD typically performing worse (Abrams et al., 2022; Dewey et al., 2007; Green et al., 2002; Kilroy et al., 2022a, 2022b;
Paquet et al., 2019). This variation in motor skill performance highlights the complexity of differentiating ASD and DCD based solely on traditional motor assessments. The discrepancy in results across studies underscores the need for more nuanced and precise diagnostic tools.
ASD is often characterized by difficulties in motor planning and execution, which are evident in the observed impairments in imitating meaningful gestures. Such motor impairments in ASD may stem from broader issues in sensorimotor integration and executive functioning, which affect the ability to understand and replicate complex movements. This is consistent with findings that individuals with ASD often struggle with tasks requiring motor dexterity, coordination, and adaptive responses to changing environmental demands (Mostofsky et al., 2006; Fournier et al., 2010).
On the other hand, children with DCD typically exhibit challenges with motor coordination, leading to difficulties in tasks that require precise control, such as handwriting or catching a ball. However, their ability to imitate gestures may not be as severely affected as in ASD. This difference can be attributed to the distinct neural pathways involved in DCD, which primarily impact motor execution rather than the integrative and planning aspects of movement (Licari et al., 2015; Fuelscher et al., 2018).
The ability to imitate gestures is a crucial aspect that differentiates ASD from DCD. While both disorders involve motor impairments, the nature and extent of these impairments differ. In ASD, the challenges extend beyond mere physical execution to encompass the understanding and planning of movements. In contrast, DCD primarily affects the mechanical execution of movement.
Furthermore, neuroimaging studies provide additional insights into these differences. For instance, individuals with ASD have shown altered activity in brain regions associated with motor planning and execution, such as the premotor cortex and cerebellum (Dapretto et al., 2006; Mostofsky et al., 2006). In contrast, studies on DCD have indicated abnormalities in regions more directly involved in motor control and coordination (Debrabant et al., 2013; Licari et al., 2015).
The emerging research using advanced technologies, such as machine learning applied to game-play data from smart tablets, offers promising avenues for more accurately distinguishing between ASD and DCD. This approach leverages the subtle nuances in motor control and planning that traditional assessments might overlook, providing a more detailed understanding of the specific motor deficits inherent in each disorder.
In summary, while traditional motor assessments show varying results in differentiating ASD and DCD, newer methods that incorporate technology and machine learning present a more accurate and nuanced approach. Understanding the specific motor challenges and their underlying neural correlates in ASD and DCD is crucial for developing targeted interventions and support strategies for each group. Continued research in this area is essential for refining diagnostic criteria and enhancing our understanding of these complex neurodevelopmental disorders.
Discussion: Differentiating ASD, DCD, and TD Using Smart Tablet-Based Motor Games
This discussion section delves into the recent findings that showcase the efficacy of using machine learning techniques applied to smart tablet-based motor games in differentiating children with Autism Spectrum Disorder (ASD), Developmental Coordination Disorder (DCD), and those typically developing (TD). These findings are particularly noteworthy given the inability of standard behavioral motor measures and video coding analysis to distinguish between ASD and DCD groups effectively.
Classifying ASD/DCD/TD by Game-Play
The study demonstrates a significant breakthrough in the application of digital technology for identifying neurodevelopmental disorders. The machine learning analysis of kinematics from smart tablet game-play categorizes ASD from TD with 76% accuracy, ASD from DCD at 71%, and DCD from TD at 78%. This achievement is unprecedented, especially considering the limitations of traditional motor assessments and visual behavioral analysis in differentiating ASD from DCD. The promise of this method in clinical diagnosis is considerable, although further research with larger sample sizes and the inclusion of social motor games in the analysis may refine and enhance the accuracy of these techniques.
Motor Markers that Distinguish Groups
The study identifies specific kinematic markers, particularly the control of deceleration and variability in gesture area, as key differentiators between the clinical groups. Autistics display more variability in gesture size compared to individuals with DCD, which could be linked to the characteristic ‘restricted and repetitive’ behaviors observed in ASD. The contrasting gesture behaviors in ASD – ranging from large, reluctant gestures to rapid, short taps – warrant further investigation to understand their distribution and implications.
Cerebellar Influence in ASD and DCD
The research extends to examining cerebellar regions, specifically crus I and II, known to be involved in sensorimotor tasks and cognitive functions like working memory, attention, and social cognition. These cerebellar regions show differential activation patterns in ASD and DCD during motor tasks, with hypoactivity observed in DCD groups during certain tasks. The distinction between ASD and DCD might lie in the cerebellar influence on motor imitation, with DCD potentially more influenced by cerebellar regions and ASD by frontal cortical areas.
Correlation Between Cerebellar Activity and Kinematic Features
Intriguingly, the study finds a correlation between activity in cerebellar regions and the kinematic features that most effectively differentiate between the groups. Specifically, activity in the left crus II correlates with Gesture Directness Variance, and the left crus I and right crus II activity correlate with Gesture Area Variance. This correlation suggests that differential activation in these cerebellar regions may underlie the motor differences observed between ASD, DCD, and TD groups.
Implications for Neurodevelopmental Disorders
These findings have profound implications for understanding the neural mechanisms underlying neurodevelopmental disorders. They suggest that while standard diagnostic tools may fail to distinguish between ASD and DCD, advanced technological methods like machine learning applied to motor games can offer a more nuanced and effective approach. The correlation between cerebellar activity and specific motor features provides a new avenue for exploring the neural basis of these disorders. Moreover, the potential for these methods to contribute to early and accurate diagnosis could significantly impact the management and treatment of ASD and DCD, leading to more tailored and effective therapeutic interventions.
In conclusion, this study marks a significant advancement in the field of neurodevelopmental disorders, offering new insights into the distinctions and similarities between ASD and DCD. The application of machine learning to smart tablet-based motor games opens up new possibilities for diagnosis and understanding of these complex conditions, with cerebellar activity playing a crucial role in differentiating between them. Future research should aim to build upon these findings, exploring the full potential of digital technology and machine learning in the diagnosis and treatment of neurodevelopmental disorders.
reference link : https://link.springer.com/article/10.1007/s10803-023-06171-8#Sec37
