Neuroimaging-Based Deep Learning in Autism Spectrum Disorder and Attention-Deficit/Hyperactivity Disorder

Area of Science:

• Neuroscience
• Artificial Intelligence
• Medical Imaging

Background:

• Neuroimaging in neurodevelopmental disorders has identified numerous biomarkers.
• Current methods struggle to integrate these biomarkers effectively for clinical use.
• Deep learning (DL) offers potential solutions for complex data analysis.

Purpose of the Study:

• To review studies combining neuroimaging and DL for neurodevelopmental disorders.
• To explore the strengths and limitations of DL in this research area.
• To identify future research directions for clinical application.

Main Methods:

• Systematic review of existing literature.
• Analysis of studies utilizing DL techniques with neuroimaging data.
• Identification of common themes, challenges, and successes.

Main Results:

• DL shows promise in analyzing complex neuroimaging data for neurodevelopmental disorders.
• Integration of multiple biomarkers via DL can enhance diagnostic and predictive power.
• Challenges remain in data standardization and clinical translation.

Conclusions:

• DL is a powerful tool for advancing neuroimaging in neurodevelopmental disorders.
• Further research is needed to optimize DL models and validate findings.
• The integration of DL holds significant potential for improving patient outcomes.