Predicting suicidal ideation from irregular and incomplete time series
of questionnaires in a smartphone-based suicide prevention platform: a
pilot study
Article Open access
Published: 06 September 2024
Gwenolé Quellec 1*, Sofian Berrouiguet 1,2 , Margot Morgiève 3,4,5,6 , Jonathan Dubois 7 ,
Marion Leboyer 8,9,10 , Guillaume Vaiva 11,12,13 , Jérôme Azé 14 & Philippe Courtet 4,7,8
1 Inserm, UMR 1101, LaTIM, IBRBS building, 22 avenue Camille Desmoulins, 29200 Brest, France.
2 Department of Psychiatry, CHU Brest, Brest, France.
3 Université Paris Cité, CNRS, Inserm, Cermes3, Paris, France.
4 Department of Emergency Psychiatry and Acute Care, Lapeyronie Hospital, CHU Montpellier, Montpellier, France.
5 ICM - Paris Brain Institute, Hôpital de la Pitié-Salpêtriére, Paris, France.
6 GEPS - Groupement d’Étude et de Prévention du Suicide, Paris, France.
7 IGF, Univ. Montpellier, CNRS, INSERM, Montpellier, France.
8 Fondation Fondamental, Hôpital Albert-Chenevier, Créteil, France.
9 Faculté de Médicine, Institut National de la Santé et de la Recherche Médicale, Université Paris-Est Créteil, Créteil, France.
10 Assistance Publique Hôpitaux de Paris, Pôle de Psychiatrie et Addictologie, Hôpitaux Universitaires Henri Mondor, Créteil, France.
11 CHU Lille, Hôpital Fontan, Department of Psychiatry, Lille, France.
12 Centre National de Resources and Résilience pour les Psychotraumatisme, Université
de Lille, Lille, France.
13 CNRS UMR-9193, SCALab - Sciences Cognitives et Sciences Affectives, Université de Lille,
Lille, France.
14 LIRMM, CNRS, Univ Montpellier, Montpellier, France. *email: gwenole.quellec@inserm.fr
Scientific Reports volume 14, Article number: 20870 (2024)
Quellec, G., Berrouiguet, S., Morgiève, M. et al. Predicting
suicidal ideation from irregular and incomplete time series of
questionnaires in a smartphone-based suicide prevention platform: a
pilot study.
Sci Rep 14, 20870 (2024). https://doi.org/10.1038/s41598-024-71760-1
Abstract
Over 700,000 people die by suicide annually. Collecting longitudinal fine-grained data about at-risk individuals, as they occur in the real world, can enhance our understanding of the temporal dynamics of suicide risk, leading to better identification of those in need of immediate intervention. Self-assessment questionnaires were collected over time from 89 at-risk individuals using the EMMA smartphone application. An artificial intelligence (AI) model was trained to assess current level of suicidal ideation (SI), an early indicator of the suicide risk, and to predict its progression in the following days. A key challenge was the unevenly spaced and incomplete nature of the time series data. To address this, the AI was built on a missing value imputation algorithm. The AI successfully distinguished high SI levels from low SI levels both on the current day (AUC = 0.804, F1 = 0.625, MCC = 0.459) and three days in advance (AUC = 0.769, F1 = 0.576, MCC = 0.386). Besides past SI levels, the most significant questions were related to psychological pain, well-being, agitation, emotional tension, and protective factors such as contacts with relatives and leisure activities. This represents a promising step towards early AI-based suicide risk prediction using a smartphone application.
Acces article https://www.nature.com/articles/s41598-024-71760-1
