Differences in the Severity and Variability of Restricted and Repetitive Behaviors in ASD Children With and Without Service Experiences

A machine learning-based diagnostic model for children with autism spectrum disorders complicated with intellectual disability

Background

Early detection of children with autism spectrum disorder (ASD) and comorbid intellectual disability (ID) can help in individualized intervention. Appropriate assessment and diagnostic tools are lacking in primary care. This study aims to explore the applicability of machine learning (ML) methods in diagnosing ASD comorbid ID compared with traditional regression models.

Method

From January 2017 to December 2021, 241 children with ASD, with an average age of 6.41 ± 1.96, diagnosed in the Developmental Behavior Department of the Children's Hospital Affiliated with the Medical College of Zhejiang University were included in the analysis. This study trained the traditional diagnostic models of Logistic regression (LR), Support Vector Machine (SVM), and two ensemble learning algorithms [Random Forest (RF) and XGBoost]. Socio-demographic and behavioral observation data were used to distinguish whether autistic children had combined ID. The hyperparameters adjustment uses grid search and 10-fold validation. The Boruta method is used to select variables. The model's performance was evaluated using discrimination, calibration, and decision curve analysis (DCA).

Result

Among 241 autistic children, 98 (40.66%) were ASD comorbid ID. The four diagnostic models can better distinguish whether autistic children are complicated with ID, and the accuracy of SVM is the highest (0.836); SVM and XGBoost have better accuracy (0.800, 0.838); LR has the best sensitivity (0.939), followed by SVM (0.952). Regarding specificity, SVM, RF, and XGBoost performed significantly higher than LR (0.355). The AUC of ML (SVM, 0.835 [95% CI: 0.747-0.944]; RF, 0.829 [95% CI: 0.738-0.920]; XGBoost, 0.845 [95% CI: 0.734-0.937]) is not different from traditional LR (0.858 [95% CI: 0.770-0.944]). Only SVM observed a good calibration degree. Regarding DCA, LR, and SVM have higher benefits in a wider threshold range.

Conclusion

Compared to the traditional regression model, ML model based on socio-demographic and behavioral observation data, especially SVM, has a better ability to distinguish whether autistic children are combined with ID.

Biopsychological correlates of repetitive and restricted behaviors in autism spectrum disorders

BACKGROUND

Autism Spectrum Disorder (ASD) is considered a neurodevelopmental condition that is characterized by alterations in social interaction and communication, as well as patterns of restrictive and repetitive behaviors (RRBs). RRBs are defined as broad behaviors that comprise stereotypies, insistence on sameness, and attachment to objects or routines. RRBs can be divided into lower-level behaviors (motor, sensory, and object-manipulation behaviors) and higher-level behaviors (restrictive interests, insistence on sameness, and repetitive language). According to the DSM-5, the grade of severity in ASD partially depends on the frequency of RRBs and their consequences for disrupting the life of patients, affecting their adaptive skills, and increasing the need for parental support.

METHODS

We conducted a systematic review to examine the biopsychological correlates of the symptomatic domains of RRBs according to the type of RRBs (lower- or higher-level). We searched for articles from the National Library of Medicine (PubMed) using the terms: autism spectrum disorders, ASD, and autism-related to executive functions, inhibitory control, inflexibility, cognitive flexibility, hyper or hypo connectivity, and behavioral approaches. For describing the pathophysiological mechanism of ASD, we also included animal models and followed PRISMA guidelines.

RESULTS

One hundred and thirty-one articles were analyzed to explain the etiology, continuance, and clinical evolution of these behaviors observed in ASD patients throughout life.

CONCLUSIONS

Biopsychological correlates involved in the origin of RRBs include alterations in a) neurotransmission system, b) brain volume, c) inadequate levels of growth factors, d) hypo- or hyper-neural connectivity, e) impairments in behavioral inhibition, cognitive flexibility, and monitoring and f) non-stimulating environments. Understanding these lower- and higher-level of RRBs can help professionals to improve or design novel therapeutic strategies.