{"title":"Social Media Images Can Predict Suicide Risk Using Interpretable Large Language-Vision Models.","authors":"Yael Badian, Yaakov Ophir, Refael Tikochinski, Nitay Calderon, Anat Brunstein Klomek, Eyal Fruchter, Roi Reichart","doi":"10.4088/JCP.23m14962","DOIUrl":null,"url":null,"abstract":"<p><p><b><i>Background:</i></b> Suicide, a leading cause of death and a major public health concern, became an even more pressing matter since the emergence of social media two decades ago and, more recently, following the hardships that characterized the COVID-19 crisis. Contemporary studies therefore aim to predict signs of suicide risk from social media using highly advanced artificial intelligence (AI) methods. Indeed, these new AI-based studies managed to break a longstanding prediction ceiling in suicidology; however, they still have principal limitations that prevent their implementation in real-life settings. These include \"black box\" methodologies, inadequate outcome measures, and scarce research on non-verbal inputs, such as images (despite their popularity today).</p><p><p><b><i>Objective:</i></b> This study aims to address these limitations and present an interpretable prediction model of clinically valid suicide risk from images.</p><p><p><b><i>Methods:</i></b> The data were extracted from a larger dataset from May through June 2018 that was used to predict suicide risk from textual postings. Specifically, the extracted data included a total of 177,220 images that were uploaded by 841 Facebook users who completed a gold-standard suicide scale. The images were represented with CLIP (Contrastive Language-Image Pre-training), a state-of-the-art deep-learning algorithm, which was utilized, unconventionally, to extract predefined interpretable features (eg, \"photo of sad people\") that served as inputs to a simple logistic regression model.</p><p><p><b><i>Results:</i></b> The results of this hybrid model that integrated theory-driven features with bottom-up methods indicated high prediction performance that surpassed common deep learning algorithms (area under the receiver operating characteristic curve [AUC] = 0.720, Cohen <i>d</i> = 0.82). Further analyses supported a theory-driven hypothesis that at-risk users would have images with increased negative emotions and decreased belongingness.</p><p><p><b><i>Conclusions:</i></b> This study provides a first proof that publicly available images can be leveraged to predict validated suicide risk. It also provides simple and flexible strategies that could enhance the development of real-life monitoring tools for suicide.</p>","PeriodicalId":50234,"journal":{"name":"Journal of Clinical Psychiatry","volume":null,"pages":null},"PeriodicalIF":4.5000,"publicationDate":"2023-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Clinical Psychiatry","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.4088/JCP.23m14962","RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"PSYCHIATRY","Score":null,"Total":0}
引用次数: 0
Abstract
Background: Suicide, a leading cause of death and a major public health concern, became an even more pressing matter since the emergence of social media two decades ago and, more recently, following the hardships that characterized the COVID-19 crisis. Contemporary studies therefore aim to predict signs of suicide risk from social media using highly advanced artificial intelligence (AI) methods. Indeed, these new AI-based studies managed to break a longstanding prediction ceiling in suicidology; however, they still have principal limitations that prevent their implementation in real-life settings. These include "black box" methodologies, inadequate outcome measures, and scarce research on non-verbal inputs, such as images (despite their popularity today).
Objective: This study aims to address these limitations and present an interpretable prediction model of clinically valid suicide risk from images.
Methods: The data were extracted from a larger dataset from May through June 2018 that was used to predict suicide risk from textual postings. Specifically, the extracted data included a total of 177,220 images that were uploaded by 841 Facebook users who completed a gold-standard suicide scale. The images were represented with CLIP (Contrastive Language-Image Pre-training), a state-of-the-art deep-learning algorithm, which was utilized, unconventionally, to extract predefined interpretable features (eg, "photo of sad people") that served as inputs to a simple logistic regression model.
Results: The results of this hybrid model that integrated theory-driven features with bottom-up methods indicated high prediction performance that surpassed common deep learning algorithms (area under the receiver operating characteristic curve [AUC] = 0.720, Cohen d = 0.82). Further analyses supported a theory-driven hypothesis that at-risk users would have images with increased negative emotions and decreased belongingness.
Conclusions: This study provides a first proof that publicly available images can be leveraged to predict validated suicide risk. It also provides simple and flexible strategies that could enhance the development of real-life monitoring tools for suicide.
期刊介绍:
For over 75 years, The Journal of Clinical Psychiatry has been a leading source of peer-reviewed articles offering the latest information on mental health topics to psychiatrists and other medical professionals.The Journal of Clinical Psychiatry is the leading psychiatric resource for clinical information and covers disorders including depression, bipolar disorder, schizophrenia, anxiety, addiction, posttraumatic stress disorder, and attention-deficit/hyperactivity disorder while exploring the newest advances in diagnosis and treatment.