Detecting Intentional Self-Harm on Instagram: Development, Testing, and Validation of an Automatic Image-Recognition Algorithm to Discover Cutting-Related Posts

Self-injurious behavior is often practiced in secrecy or involves body parts that are easy to hide, making early detection difficult and hampering intervention and treatment. However, cutting, one of the most common intentional forms of nonsuicidal self-injury (NSSI), is relatively often shared publicly via new digital media technologies. We explored NSSI on Instagram through a pioneering combination of two computational methods: First, we developed an automatic image-recognition algorithm that uncovered NSSI (or the absence of NSSI) in digital pictures, and second, we employed web-scraping techniques to obtain all pictures posted on Instagram in a given time frame under four NSSI-related hashtags in English and German. The image-recognition algorithm was then used to explore the relative prevalence of NSSI in these N = 13,132 pictures posted within 48 hr on Instagram under #cutting (n = 4,219), #suicide (n = 7,910), #selbstmord (n = 173), and #ritzen (n = 830) in June 2018. This article not only aims to raise awareness of NSSI on Instagram but also introduces the first automatic image-recognition algorithm that addresses cutting on social media and presents that algorithm’s first empirical test run on a large sample of pictures scraped from Instagram. The ultimate goal of this research is to protect vulnerable populations from contact with NSSI-related pictures posted on social media.

[1]  P. Hasking,et al.  The relationship between portrayals of nonsuicidal self-injury, attitudes, knowledge, and behavior. , 2013, Crisis.

[2]  A. Bruns,et al.  Twitter hashtags from ad hoc to calculated publics , 2015 .

[3]  K. Hawton,et al.  Predicting deliberate self-harm in adolescents: a six month prospective study. , 2009, Suicide & life-threatening behavior.

[4]  Andrea Vedaldi,et al.  MatConvNet: Convolutional Neural Networks for MATLAB , 2014, ACM Multimedia.

[5]  Mario Haim,et al.  Equal access to online information? Google’s suicide-prevention disparities may amplify a global digital divide , 2018, New Media Soc..

[6]  I. Ajzen.  The theory of planned behavior , 1991 .

[7]  M. Slater.  Reinforcing Spirals: The Mutual Influence of Media Selectivity and Media Effects and Their Impact on Individual Behavior and Social Identity , 2007 .

[8]  ImageNet Classification with Deep Convolutional Neural , 2013 .

[9]  Ellen M. Selkie,et al.  Secret Society 123: Understanding the Language of Self-Harm on Instagram. , 2016, The Journal of adolescent health : official publication of the Society for Adolescent Medicine.

[10]  Debbie Ging,et al.  ‘Written in these scars are the stories I can’t explain’: A content analysis of pro-ana and thinspiration image sharing on Instagram , 2018, New Media Soc..

[11]  P. Hasking,et al.  Cutting Words: A Commentary on Language and Stigma in the Context of Nonsuicidal Self-Injury. , 2018, The Journal of nervous and mental disease.

[12]  Xiaolong Wang,et al.  Topic sentiment analysis in twitter: a graph-based hashtag sentiment classification approach , 2011, CIKM '11.

[13]  G. Lewis,et al.  Differences in risk factors for self-harm with and without suicidal intent: Findings from the ALSPAC cohort , 2014, Journal of affective disorders.

[14]  R. Evans,et al.  The influence of online images on self-harm: A qualitative study of young people aged 16–24 , 2017, Journal of adolescence.

[15]  Stephen P. Lewis,et al.  A Double-Edged Sword: A Review of Benefits and Risks of Online Nonsuicidal Self-Injury Activities. , 2016, Journal of clinical psychology.

[16]  L. Hartling,et al.  A Systematic Review of Social Media Use to Discuss and View Deliberate Self-Harm Acts , 2016, PloS one.

[17]  Michele L. Ybarra,et al.  Exposure to websites that encourage self-harm and suicide: prevalence rates and association with actual thoughts of self-harm and thoughts of suicide in the United States. , 2014, Journal of adolescence.

[18]  Steven Stack,et al.  Changes in suicide rates following media reports on celebrity suicide: a meta-analysis , 2012, Journal of Epidemiology & Community Health.

[19]  M. Slater.  Reinforcing Spirals Model: Conceptualizing the Relationship Between Media Content Exposure and the Development and Maintenance of Attitudes , 2015, Media psychology.

[20]  F. Arendt.  Suicide on Instagram – Content Analysis of a German Suicide-Related Hashtag , 2019, Crisis.

[21]  Jonathan Berant,et al.  Polyglot Semantic Parsing in APIs , 2018, NAACL-HLT.

[22]  Carsten Reinemann,et al.  First do no harm: Cross-sectional and longitudinal evidence for the impact of individual suicidality on the use of online health forums and support groups , 2016, Comput. Hum. Behav..

[23]  Cornelia Caragea,et al.  Content-Driven Detection of Cyberbullying on the Instagram Social Network , 2016, IJCAI.

[24]  Stephen P. Lewis,et al.  Non-Suicidal Self-Injury in Our Schools: A Review and Research-Informed Guidelines for School Mental Health Professionals , 2017 .

[25]  T. Willoughby,et al.  Examining the link between nonsuicidal self-injury and suicidal behavior: a review of the literature and an integrated model. , 2012, Clinical psychology review.

[26]  M. Nock,et al.  Nonsuicidal self-injury: Definition and classification. , 2009 .

[27]  M. Prinstein,et al.  A functional approach to the assessment of self-mutilative behavior. , 2004, Journal of consulting and clinical psychology.

[28]  Robert A. Fahey,et al.  Tracking the Werther Effect on social media: Emotional responses to prominent suicide deaths on twitter and subsequent increases in suicide. , 2018, Social science & medicine.

[29]  S. Lewis,et al.  Responses to Online Photographs of Non-Suicidal Self-Injury: A Thematic Analysis , 2013, Archives of suicide research : official journal of the International Academy for Suicide Research.

[30]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[31]  D. Phillips.  The influence of suggestion on suicide: substantive and theoretical implications of the Werther effect. , 1974, American sociological review.

[32]  P. Räsänen,et al.  Victimization and Exposure to Pro‐Self‐Harm and Pro‐Suicide Websites: A Cross‐National Study , 2017, Suicide & life-threatening behavior.

[33]  I. Ajzen,et al.  Attitudinal and normative variables as predictors of specific behavior. , 1973 .

[34]  Jonathan S. Comer,et al.  Examining the scope and patterns of deliberate self‐injurious cutting content in popular social media , 2017, Depression and anxiety.

[35]  T. Niederkrotenthaler,et al.  The role of language in suicide reporting: Investigating the influence of problematic suicide referents. , 2018, Social science & medicine.

[36]  Axel Bruns,et al.  Towards a typology of hashtag publics: a large-scale comparative study of user engagement across trending topics , 2016 .

[37]  Michael S. Bernstein,et al.  ImageNet Large Scale Visual Recognition Challenge , 2014, International Journal of Computer Vision.

[38]  F. Keller,et al.  #cutting: Non-suicidal self-injury (NSSI) on Instagram , 2017, Psychological Medicine.