Surveillance of Suicide Ideation in Adolescents (VISIA)

Обзор

The goal of this observational study is to validate an AI algorithm's capability to differentiate the population with suicidal ideation from a control population using various multimodal variables, including voice analysis, facial emotions, natural language, and proteomics data. The primary research question it aims to answer is: Is it possible to identify suicidal ideation and suicide risk in adolescents early and non-intrusively using multimodal data analysis through digital instruments equipped with artificial intelligence? Participants in this study will be asked to: Complete psychometric instruments to establish a gold standard for detecting suicide risk and suicidal ideation. Provide voice recordings, facial emotion data, and linguistic content in natural and specific contexts. Participate in salivary proteomics data collection. This study compares three distinct groups: Ideation: Adolescent patients with current suicidal ideation. Clinical Population: Psychological or psychiatric patients of the same age and gender without suicidal ideation. General Population: Adolescents without known psychological or psychiatric pathology of the same age and gender, without suicidal ideation. Researchers will compare these groups to determine if the AI algorithm is effective in differentiating individuals with suicidal ideation (Group 1) from both a clinical control group (Group 2) and a general population control group (Group 3) using the collected multimodal data. The study aims to assess the algorithm's ability to identify early signs of suicide risk in these distinct participant populations.

Подробное описание

Introduction:

The Surveillance of Suicide Ideation in Adolescents (VISIA) project is a collaborative effort involving a diverse research team encompassing experts in mental disorders and seasoned researchers in data science, specializing in intelligent services and applications across various domains. The project aims to address the shortcomings identified in existing research. The initial phase of this endeavor seeks to validate an artificial intelligence (AI) algorithm capable of distinguishing individuals with suicidal ideation from a control group by analyzing multimodal variables. These variables include voice analysis, facial emotions, natural language, and proteomics. The cases will be labeled following established protocols.

Hypothesis:

The primary research question driving this project is: Can early and non-intrusive identification of suicidal ideation and suicide risk in young adults be achieved using multimodal data analysis through digital instruments equipped with artificial intelligence?

Objectives:

The overarching objective of this proposal is to design and implement an automated system for detecting emotional distress and suicide risk in adolescents aged between 11 and 16 using AI. This detection relies on molecular profiles derived from salivary proteomics, video and audio recordings in diverse contexts, and the linguistic content generated by the study participants. The integration of molecular profiles with data from video and audio recordings promises enhanced detection capabilities, providing a competitive edge in the development of future diagnostic tools.

Specific Objectives:

To realize the goals outlined above and tackle the scientific and technological challenges, the following specific objectives are established:

Development of a gold standard for detecting suicide risk and suicidal ideation in adolescents.

Identification of biomarkers (e.g., voice, facial expressions, salivary molecular fingerprint) relevant to characterizing adolescents with regard to suicide risk.

Identification of pertinent multimodal features for characterizing adolescents in relation to suicide risk.

Labeling of samples and AI model training.

Methodology:

Design and Study Subjects:

The study adopts a non-interventionist, analytical, observational, and prospective approach. It encompasses three distinct groups for comparison:

Ideation: Comprising individuals with current suicidal ideation. Clinical Population: Consisting of psychological or psychiatric patients of the same age and gender, devoid of suicidal ideation.

General Population: Encompassing adolescents with no known psychological or psychiatric issues, matched in age and gender, and free of suicidal ideation.

All three groups will undergo assessments using psychometric instruments, serving as a gold standard for AI system training and validation.

In summary, the VISIA project endeavors to leverage artificial intelligence and multimodal data analysis to detect suicidal ideation and suicide risk in adolescents, offering the potential for early and non-intrusive identification. The study's comprehensive methodology, involving diverse subject groups and rigorous data collection, is poised to contribute significantly to the field of mental health research and diagnostic tool development.

Первичные конечные точки

• Strengths and Difficulties Questionnaire (SDQ) [Срок оценки: through study completion, an average of 1 year]
• Mood and Feelings Questionnaire (MFQ) [Срок оценки: through study completion, an average of 1 year]
• Paykel Suicide Scale (PSS) [Срок оценки: through study completion, an average of 1 year]
• Facial microexpressions, voice and linguistic content analysis [Срок оценки: through study completion, an average of 1 year]
• Proteomic and stress level analysis [Срок оценки: through study completion, an average of 1 year]

Вторичные конечные точки (3)

• Millon Adolescent Clinical Inventory - II (MACI-II) [Срок оценки: through study completion, an average of 1 year]
• Problematic Internet Use Scale in Adolescents (EUPI-A) [Срок оценки: through study completion, an average of 1 year]
• European Bullying Intervention Project Questionnaire (EBIP-Q) and European Ciberbullying Intervention Project Questionnaire (ECIP-Q) [Срок оценки: through study completion, an average of 1 year]

Критерии участия

Критерии включения

Ideation Group

• Participants must be between 11 and 16 years old.
• Patients who do not meet any of the exclusion criteria and whose parents (both), guardians, or legal representatives accept their voluntary participation in the study. In case of divorce, authorization from both parents will be required.
• Patients receive care in child and adolescent psychiatry or psychology services and currently have suicidal ideation.
• Patients may have a previous follow-up, as long as they have suicidal ideation at the present moment, or they may be new consultations and may be receiving pharmacological treatment (which will be documented in the study).

Clinical Population Group

• Participants must be between 11 and 16 years old.
• Patients who do not meet any of the exclusion criteria and whose parents (both), guardians, or legal representatives accept their voluntary participation in the study. In case of divorce, authorization from both parents will be required.
• Patients receive care in child and adolescent psychiatry or psychology services for any reason other than those listed as exclusion criteria.
• Patients have not had suicidal ideation in the past year. Patients may have a previous follow-up or may be new consultations, and they may be receiving pharmacological treatment (which will be documented in the study).

General Population Group

• Participants must be between 11 and 16 years old.
• Students who do not meet any of the exclusion criteria and whose parents (both), guardians, or legal representatives accept their voluntary participation in the study. In case of divorce, authorization from both parents will be required.
• Students must declare that they have not received psychiatric or psychological follow-up before, nor have they had suicidal ideation in the past year.

Exclusion Criteria (applicable to all groups):

• Neurodevelopmental disorders or speech disorders affecting verbal communication capacity, specifically including:
• Autism Spectrum Disorder grades 2 and 3.
• Communication disorders (selective mutism, language disorder, phonological disorder, childhood-onset fluency disorder, social communication disorder).
• Individuals with hearing impairment and verbal language impairment.
• Individuals with moderate and severe intellectual disabilities.
• Individuals with psychotic disorders.

Критерии приведены из реестра в оригинале (на английском). Окончательную оценку соответствия проводит исследовательский центр.

Здоровые добровольцы: Да

Дизайн исследования

Модель наблюдения

Случай-контроль

Центры проведения

Испания · 4 центра

• Servizo Galego de Saúde — Santiago de Compostela
• Hospital Álvaro Cunqueiro — Vigo
• University of Vigo — Vigo
• University Hospital Complex of Ourense — Ourense

Публикации

• Abd-Alrazaq AA, Alajlani M, Alalwan AA, Bewick BM, Gardner P, Househ M. An overview of the features of chatbots in mental health: A scoping review. Int J Med Inform. 2019 Dec;132:103978. doi: 10.1016/j.ijmedinf.2019.103978. Epub 2019 Sep 25. PMID 31622850
• Adamou M, Antoniou G, Greasidou E, Lagani V, Charonyktakis P, Tsamardinos I, Doyle M. Toward Automatic Risk Assessment to Support Suicide Prevention. Crisis. 2019 Jul;40(4):249-256. doi: 10.1027/0227-5910/a000561. Epub 2018 Nov 26. PMID 30474411
• Bachmann S. Epidemiology of Suicide and the Psychiatric Perspective. Int J Environ Res Public Health. 2018 Jul 6;15(7):1425. doi: 10.3390/ijerph15071425. PMID 29986446
• Briere J, Gil E. Self-mutilation in clinical and general population samples: prevalence, correlates, and functions. Am J Orthopsychiatry. 1998 Oct;68(4):609-20. doi: 10.1037/h0080369. PMID 9809120
• Brunner R, Kaess M, Parzer P, Fischer G, Carli V, Hoven CW, Wasserman C, Sarchiapone M, Resch F, Apter A, Balazs J, Barzilay S, Bobes J, Corcoran P, Cosmanm D, Haring C, Iosuec M, Kahn JP, Keeley H, Meszaros G, Nemes B, Podlogar T, Postuvan V, Saiz PA, Sisask M, Tubiana A, Varnik A, Wasserman D. Life-time prevalence and psychosocial correlates of adolescent direct self-injurious behavior: a compar PMID 24215434
• Choi D, Sumner SA, Holland KM, Draper J, Murphy S, Bowen DA, Zwald M, Wang J, Law R, Taylor J, Konjeti C, De Choudhury M. Development of a Machine Learning Model Using Multiple, Heterogeneous Data Sources to Estimate Weekly US Suicide Fatalities. JAMA Netw Open. 2020 Dec 1;3(12):e2030932. doi: 10.1001/jamanetworkopen.2020.30932. PMID 33355678
• Krageloh CU, Czuba KJ, Billington DR, Kersten P, Siegert RJ. Using feedback from patient-reported outcome measures in mental health services: a scoping study and typology. Psychiatr Serv. 2015 Mar 1;66(3):224-41. doi: 10.1176/appi.ps.201400141. Epub 2014 Dec 1. PMID 25727110
• Gillies D, Christou MA, Dixon AC, Featherston OJ, Rapti I, Garcia-Anguita A, Villasis-Keever M, Reebye P, Christou E, Al Kabir N, Christou PA. Prevalence and Characteristics of Self-Harm in Adolescents: Meta-Analyses of Community-Based Studies 1990-2015. J Am Acad Child Adolesc Psychiatry. 2018 Oct;57(10):733-741. doi: 10.1016/j.jaac.2018.06.018. Epub 2018 Aug 21. PMID 30274648

Идентификаторы

NCT: NCT06341634 · VISIA · TED2021-130747B-C22 · 3494130747-130747-28-521

Первоисточники (государственные реестры)

• ClinicalTrials.gov (NIH, США)

Открыть это исследование на ClinicalTrials.gov ↗