Instagram photos reveal predictive markers of depression

Abstract

Using Instagram data from 166 individuals, we applied machine learning tools to successfully identify markers of depression. Statistical features were computationally extracted from 43,950 participant Instagram photos, using color analysis, metadata components, and algorithmic face detection. Resulting models outperformed general practitioners' average unassisted diagnostic success rate for depression. These results held even when the analysis was restricted to posts made before depressed individuals were first diagnosed. Human ratings of photo attributes (happy, sad, etc.) were weaker predictors of depression, and were uncorrelated with computationally-generated features. These results suggest new avenues for early screening and detection of mental illness.

Recreating the Relationship between Subjective Wellbeing and Personality Using Machine Learning: An Investigation into Facebook Online Behaviours

“…For example, most users on Twitter were classified as emotionally stable and extroverted by using counts of the Twitter information 'following', 'followers', and 'listed' [28]. Additionally, Reece and Danforth [29] successfully identified markers of depression from participants' Instagram photos, which surpassed general practitioners' typical unassisted diagnostic success rate for depression. Quercia, et al [30] evaluated the efficacy of digital methods to predict the strength of online social relationships on Facebook and their findings were consistent with previous analyses used on Twitter [31].…”

Exploring the Utility of Community-Generated Social Media Content for Detecting Depression: An Analytical Study on Instagram (Preprint)

“…Much of the previous research has largely been focused on understanding a user's mental well-being through information that the user posts on social media, such as Twitter "Tweets", Facebook status updates, or Instagram images [10][11][12]17]. Other studies, however, have suggested that community-generated data is correlated with user-generated data, as alcohol-related posts have more positive community-generated data [24].…”

“…Prediction of MDD based on social media data is well established with strong results [10,11,17,22,28]. However, the existing literature has largely focused on using only user-generated data for this purpose.…”

“…Instagram posts consist of a photo or video, and, optionally, a caption, and comment and/or "likes" from other users. Instagram photos have been shown to contain information relevant to depression status [10].…”

“…Recent evidence shows significant predictive power to identify MDD, particularly in users of Twitter and Facebook [10][11][12][13]. However, the majority of research conducted to predict MDD in social media users has focused on user-generated content, i.e., the content created by the user themselves.…”