AI for Childhood Neurodevelopmental Disorder Diagnosis: An Umbrella Review

An umbrella review synthesizing evidence from 64 systematic reviews and meta-analyses on the use of artificial intelligence for diagnosing neurodevelopmental disorders in children. The review, authored by Alejandro Alberca-González and published on figshare in March 2026, covers studies from 2015 to 2025, focusing on conditions like autism spectrum disorder and attention-deficit/hyperactivity disorder. It analyzes AI model performance across data modalities including neuroimaging, electrophysiological, and clinical data.

Use Cases

• Benchmarking AI diagnostic accuracy for autism spectrum disorder based on reported performance metrics.
• Identifying prevalent data modalities like neuroimaging or electrophysiological signals used in AI models for ADHD diagnosis.
• Assessing methodological quality trends in AI research for neurodevelopmental disorders using the AMSTAR-2 criteria mentioned.
• Comparing the application of specific AI models like support vector machines or convolutional neural networks across different disorders.

Strengths

• Synthesizes evidence from 64 studies, providing a broad overview of the field.
• Reports specific diagnostic accuracy ranges from 66% to 99% across different data modalities.
• Assesses methodological quality, noting 80% of studies were rated critically low and 5% high.
• Covers a defined time range of studies published from January 2015 to August 2025.

Limitations

• The dataset is a 139.2 KB DOCX file containing review text, not the primary clinical or imaging data.
• Row count and column-level documentation for any underlying data are unknown.
• The review notes current research limitations like insufficient external validation and lack of standardization.

Provenance

Source

figshare, author Alejandro Alberca-González.

Collection Method

Systematic umbrella review conducted via searches on Web of Science, PsycINFO, and PubMed.

Time Range

Studies published from January 2015 to August 2025.

Freshness

Last updated 2026-03-18 06:58:02.