Has Machine Learning Enhanced the Diagnosis of Autism Spectrum Disorder?

Area of Science:

• Neuroscience
• Developmental Psychology
• Computer Science

Background:

• Autism Spectrum Disorder (ASD) is a complex developmental disorder impacting communication and learning.
• Early diagnosis is crucial for effective rehabilitation and social integration.
• Traditional ASD assessment methods have limitations in real-world accuracy.

Purpose of the Study:

• To review recent technological advancements in ASD diagnosis.
• To analyze the impact of these technologies on individuals with ASD and their families.
• To explore the application of machine learning in identifying ASD.

Main Methods:

• Review of current research on technology-assisted ASD diagnosis.
• Analysis of unconventional data sources (electrophysiological, virtual reality, eye-tracking).
• Examination of machine learning algorithms (LSTM, CNN, RF, NB) for ASD identification.

Main Results:

• Technological advancements enable processing of novel data for ASD assessment.
• Virtual reality, eye-tracking, and emotion recognition studies show promise.
• Machine learning techniques are increasingly utilized for discriminating ASD from typical development.

Conclusions:

• Innovative diagnostic approaches are enhancing ASD identification.
• Further research is needed to refine methodologies for diverse ASD presentations.
• Improved diagnostic tools can lead to better life quality for individuals with ASD.