EPP405
Use of Artificial Intelligence and mobile applications in the early detection of suicide risk among young people across the globe
M. C. B. Maia 1,*, B. E. Bandeira 1, L. A. Brito 1
1Universidade de Brasilia, Brasilia, Brazil
Introduction: Youth suicide is a major global public health issue, the fourth leading cause of death for ages 15–19, according to graphic 1 (Hughes et al. J Adolesc Health 2023; 72:367–375). Incidence rises from age 10, especially in males (3:1 ratio). Early detection is difficult due to multifactorial causes; depressive symptoms and subtle behavioral changes are often missed. With 85% global wireless access, mobile apps and AI offer continuous monitoring and personalized interventions. This study reviews their contributions and limitations, including mood monitoring, voice analysis, and social media analysis, evaluating usability, effectiveness, and ethical challenges.
Objectives: This study evaluates mobile apps and AI for early suicide risk detection in youth, reviewing tools like mood monitoring, voice analysis, geolocation, and social media. It assesses effectiveness, usability, and ethical challenges, highlighting their potential to prevent suicide and support mental health.
Methods: A literature review (since 2017) used PubMed, SciELO, BMC Psychiatry, SpringerLink, JMIR, and Arxiv. Keywords included ‘suicide prevention’, ‘AI’, ‘mobile health applications’, ‘youth’, and ‘digital phenotyping’. Included studies addressed mobile apps or AI for adolescent/young adult suicide risk; studies outside the population or without digital tools were excluded.
Results: Apps like BeyondNow (Australia), BlueIce (UK), LifeBuoy (Australia), and eRAPPORT (France) show good usability and real-time mood monitoring (Larsen et al. PLoS One 2016; 11:e0152285). Evidence for reducing suicidal ideation exists, but prevention of attempts remains limited (Baldessarini et al. Harv Rev Psychiatry 2020; 28:15–29). As shown on graphic 2 (Witt et al. Lancet Psychiatry 2017; 4:486–500; Torous et al. Lancet Digit Health 2021; 3:e555–e562), AI using passive data (GPS, time at home, voice) and digital phenotyping achieves promising predictive accuracy Latin America (Chile, Brazil, Colombia) adapts global tools or develops local solutions for remote monitoring. According to table 1 (Souza et al. Rev Bras Psiquiatr 2021; 43:609–617; WHO 2022), Europe conducts the majority of robust studies and clinical trials, whereas Latin America prioritises local projects and pilot innovations, grappling with structural challenges such as digital inequality and the necessity of public policies for integration.
Image 1:
Image 2:
Image 3:
Conclusions: Mobile apps and AI enable early detection, continuous monitoring, and personalized interventions, but challenges remain: validation, healthcare integration, data protection, and ethics. Hybrid models combining human care and digital tools are key. Europe leads in AI innovation; Latin America focuses on accessible, locally adapted solutions amid digital inequalities and policy needs.
Disclosure of Interest: None Declared