Artificial Intelligence Based Techniques for the Detection of Socio-Behavioral Disorders: A Systematic Review

In new era of emerging technologies, diagnosing the neurodevelopmental disorders at the primitive stage has led to a rise in the development of automated diagnostic systems. Socio-behavioral disorder (SBD),a subtype of NDDs, being amongst the most complex behavioral impairments is the root cause for disability, specifically in young children. Under the term SBD, falls a group of heterogeneous disorders; Autism Spectrum Disorder (ASD) and Attention Deficit Hyperactivity Disorder (ADHD) whose diagnosis remains a challenging task due to the spectrum of conditions associated to it and also due to their high comorbidity with each other. Traditional based recognition of SBD is based on qualitative parameters which make diagnosis more time consuming. Hence, diagnosis of SBD is a preeminent area that demands deliberations from the researchers, scientists as well as academicians to introduce more automated diagnostic systems based on quantitative parameters that not only provides reliability and accuracy besides affordable for everyone. Motivated by these facts,this article seeks to provide a thorough investigation of efforts on the subject of automated diagnostic systems using artificial intelligence techniques based on quantitative parameters (biomarkers and multimodal data) for early ASD or ADHD detection. An extensive survey of AI based diagnostic systems for ASD, ADHD are also presented. Furthermore, the studies of various automated AI based diagnostic systems for the detection of ASD, ADHD and ASD+ADHD are presented in this research work. Finally, this article highlights the open issues existing in the literature that needs to be explored further for providing more effective SBD analysis and also presented an outline of the proposed work which will be an aid in the diagnostic field of ASD, ADHD and comorbid ASD.

Automated summary

The article investigates the development of automated diagnostic systems based on quantitative parameters for early detection of ASD or ADHD, presenting a survey of AI-based diagnostic systems for ASD and ADHD. Additionally, studies of various automated AI-based diagnostic systems for ASD, ADHD and comorbid ASD are discussed, highlighting open issues in the literature that need further exploration.