Machine Learning Applications in Mental Health and Substance Use Research Among the LGBTQ2S+ Population: Scoping Review (Preprint)

Abstract

BACKGROUND People at high risk of mental health or substance addiction issues among sexual and gender minorities may have more nuanced characteristics that may not be easily discovered by traditional statistical methods. OBJECTIVE This review aimed at identifying literature that used machine learning to investigate mental health or substance use concerns among lesbian, gay, bisexual, transgender, queer or questioning and two-spirit (LGBTQ2S+) population as well as directing future research in this field. METHODS MEDLINE, EMBASE, PubMed, CINAHL Plus, PsycINFO and IEEE Xplore, Summon databases were searched from November to December 2020. We included original studies which used machine learning to explore mental health and/or substance use among LGBTQ2S+ population and excluded studies of genomics and pharmacokinetics. Two independent reviewers reviewed all papers and extracted data on general study findings, model development and discussion of study findings. RESULTS We included 11 studies in this review, of which 9 (82%) studies were on mental health and only 2 (18%) studies were on substance use concerns. All studies were published within last 2 years and majority were conducted in the Unites States. Among mutually non-exclusive population categories, sexual minorities male were the most commonly studied subgroup (n=5, 45%), while sexual minorities female were studied the least (n=2, 18%). Studies were categorized into 3 major domains: online content analysis (n=6, 55%), prediction modelling (n=4, 36%) and imaging study (n=1, 9%). CONCLUSIONS Machine learning can be a promising tool of capturing and analyzing hidden data of mental health and substance use concerns among LGBTQ2S+ people. In addition to conducting more research on sexual minority women, different mental health and substance use problems as well as outcomes, future research should explore newer environments and data sources and intersections with various social determinants of health.

Machine Learning Algorithms for Predicting Factitious Disorder Using the Learning Vector Quantization Method

“…Mental to Childrearing Fathers (MSCF) can be an effective tool for mental health screening among fathers and children [4].…”