The Role of AI-Driven Tools for Early Detection of Mental Health Disorders: A Systematic Review

Abstract

The objective of this study was to systematically review the use of AI and related tools in the detection of mental health disorders. A systematic literature review was carried out on the studies searched and gathered from 1st June 2024 to 30th August 2024. The sample included 12 research papers addressing a vast array of AI applications and neurological conditions; furthermore, SPSS and Excel software were used for performing statistical analysis and producing relevant graphs on the aforementioned chosen studies. The study included mental disorders in broad terms (18.9%), Depression (13.5%), Anxiety (10.8%), Schizophrenia (8.1%), autism spectrum disorder (8.1%), Epilepsy (5.4%), Bipolar (5.4%), PTSD (2.7%), Intellectual Disability (2.7%), Identity Disorder (2.7%), Suicide (2,7%), Self-Harm (2.7%), ADHD (2.7%), Dyslexia (2.7%), Tourette's Syndrome (2.7%), Obsessive Compulsive disorder (2.7%), Mental Disorders of visual field (2.7%), and Substance Abuse Disorder (2.7%). About 66.6% of the studies containing information about their AI application’s efficacy rated it moderate or higher than 50% with a further 22.2% of studies containing mixed results, overall suggesting a high positive correlation between the use of AI in detecting mental health disorders with appropriate human oversight required in addressing said conditions.

Keywords: artificial intelligence, psychopathology, meta-analysis, mental health