Artificial Intelligence and Machine Learning Models Examined in Early Autism Diagnosis
Machine learning models developed for the early detection of autism spectrum disorder were evaluated through a systematic review with a pediatric and engineering perspective.
A comprehensive systematic review examining the role of artificial intelligence and machine learning models in the early diagnosis of autism spectrum disorder has been published, addressing multiple data modalities and diagnostic performances within the scope of the research.
Autism Spectrum Disorder and the Importance of Early Diagnosis
Autism spectrum disorder is a complex neurodevelopmental condition characterized by persistent differences in social communication and repetitive behavioral patterns.
The wide variation in clinical presentation among individuals makes early diagnosis difficult and increases the need for accessible, reliable assessment pathways.
Artificial Intelligence and Machine Learning Approaches
Artificial intelligence technologies are applied to various data sources for autism diagnosis by modeling non-linear relationships in high-dimensional data.
Common models such as support vector machines, random forests, and deep neural networks support classification processes.
Research Objectives and Scope
The main objectives of the study include systematically summarizing methods related to early autism detection and evaluating diagnostic performances.
Additionally, considerations regarding engineering readiness, external validation, explainability, and clinical application were also examined.
Data Sources and Search Strategy
Within the scope of the research, PubMed and IEEE Xplore databases were searched, and peer-reviewed original studies and conference proceedings meeting specific criteria were examined.
Through phased filterings among thousands of records obtained from the initial screening results, twenty-one primary studies were selected for the final synthesis.