Instagram photos reveal predictive markers of depression

Springer Science and Business Media LLC - Tập 6 - Trang 1-12 - 2017
Andrew G Reece1, Christopher M Danforth2,3,4
1Department of Psychology, Harvard University, Cambridge, USA
2Computational Story Lab, Vermont Advanced Computing Core, University of Vermont, Burlington, USA
3Department of Mathematics and Statistics, University of Vermont, Burlington, USA
4Vermont Complex Systems Center, University of Vermont, Burlington, USA

Tóm tắt

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.

Tài liệu tham khảo

Moreno M, Christakis D, Egan K, Brockman L, Becker T (2012) Associations between displayed alcohol references on Facebook and problem drinking among college students. Arch Pediatr Adolesc Med 166(2):157-163. doi:10.1001/archpediatrics.2011.180 De Choudhury M, Counts S, Horvitz E (2013) Predicting postpartum changes in emotion and behavior via social media. In: Proceedings of the SIGCHI conference on human factors in computing systems. ACM, New York, pp 3267-3276. doi:10.1145/2470654.2466447 De Choudhury M, Counts S, Horvitz EJ, Hoff A (2014) Characterizing and predicting postpartum depression from shared Facebook data. In: Proceedings of the 17th ACM conference on computer supported cooperative work & social computing, ACM, New York, pp 626-638. doi:10.1145/2531602.2531675 De Choudhury M, Gamon M, Counts S, Horvitz E (2013) Predicting depression via social media. In: Seventh international AAAI conference on weblogs and social media Katikalapudi R, Chellappan S, Montgomery F, Wunsch D, Lutzen K (2012) Associating Internet usage with depressive behavior among college students. IEEE Technol Soc Mag 31(4):73-80. doi:10.1109/MTS.2012.2225462 Moreno MA, Jelenchick LA, Egan KG, Cox E, Young H, Gannon KE, Becker T (2011) Feeling bad on Facebook: depression disclosures by college students on a social networking site. Depress Anxiety 28(6):447-455. doi:10.1002/da.20805 Coppersmith G, Harman C, Dredze M (2014) Measuring post traumatic stress disorder in Twitter. In: Eighth international AAAI conference on weblogs and social media De Choudhury M, Kiciman E, Dredze M, Coppersmith G, Kumar M (2016) Discovering shifts to suicidal ideation from mental health content in social media. In: Proceedings of the 2016 CHI conference on human factors in computing systems. ACM, New York, pp 2098-2110. doi:10.1145/2858036.2858207 Christakis NA, Fowler JH (2010) Social network sensors for early detection of contagious outbreaks. PLoS ONE 5(9):e12948. doi:10.1371/journal.pone.0012948 Schmidt CW (2012) Trending now: using social media to predict and track disease outbreaks. Environ Health Perspect 120(1):a30-a33. doi:10.1289/ehp.120-a30 Paparrizos J, White RW, Horvitz E (2016) Screening for pancreatic adenocarcinoma using signals from web search logs: feasibility study and results J Oncol Pract 12(8):737-744. doi:10.1200/JOP.2015.010504 Instagram (2016) Instagram press release. Available at https://www.instagram.com/press/. Accessed July 26, 2016 Chaffey D (2016) Global social media research summary 2016. Available at bit.ly/1WRviEI. Accessed July 19, 2016 Lup K, Trub L, Rosenthal L (2015) Instagram #Instasad?: exploring associations among Instagram use, depressive symptoms, negative social comparison, and strangers followed. Cyberpsychol Behav 18(5):247-252. doi:10.1089/cyber.2014.0560 Andalibi N, Ozturk P, Forte A (2015) Depression-related imagery on Instagram. In: Proceedings of the 18th ACM conference companion on computer supported cooperative work & social computing, ACM, New York, pp 231-234. doi:10.1145/2685553.2699014 Boyatzis CJ, Varghese R (1994) Children’s emotional associations with colors. J Genet Psychol 155(1):77-85 Carruthers HR, Morris J, Tarrier N, Whorwell PJ (2010) The Manchester Color Wheel: development of a novel way of identifying color choice and its validation in healthy, anxious and depressed individuals. BMC Med Res Methodol 10:12. doi:10.1186/1471-2288-10-12 Hemphill M (1996) A note on adults’ color-emotion associations. J Genet Psychol 157(3):275-280 Barrick CB, Taylor D, Correa EI (2002) Color sensitivity and mood disorders: biology or metaphor? J Affect Disord 68(1):67-71. doi:10.1016/S0165-0327(00)00358-X American Psychiatric Association (2000) Diagnostic and statistical manual of mental disorders, 4th edn. doi:10.1176/appi.books.9780890423349 Bruce ML, Hoff RA (1994) Social and physical health risk factors for first-onset major depressive disorder in a community sample. Soc Psychiatry Psychiatr Epidemiol 29(4):165-171. doi:10.1007/BF00802013 Cornford CS, Hill A, Reilly J (2007) How patients with depressive symptoms view their condition: a qualitative study. Fam Pract 24(4):358-364. doi:10.1093/fampra/cmm032 Karp DA (1994) Living with depression: illness and identity turning points. Qual Health Res 4(1):6-30. doi:10.1177/104973239400400102 Mitchell AJ, Vaze A, Rao S (2009) Clinical diagnosis of depression in primary care: a meta-analysis. Lancet 374(9690):609-619. doi:10.1016/S0140-6736(09)60879-5 Epstein RM, Duberstein PR, Feldman MD, Rochlen AB, Bell RA, Kravitz RL et al. (2010) ‘I didn’t know what was wrong:’ how people with undiagnosed depression recognize, name and explain their distress. J Gen Intern Med 25(9):954-961. doi:10.1007/s11606-010-1367-0 Radloff LS (1977) The CES-D scale: a self-report depression scale for research in the general population. Appl Psych Manage 1(3):385-401. doi:10.1177/014662167700100306 Fountoulakis KN, Bech P, Panagiotidis P, Siamouli M, Kantartzis S, Papadopoulou A et al. (2007) Comparison of depressive indices: reliability, validity, relationship to anxiety and personality and the role of age and life events. J Affect Disord 97(1-3):187-195. doi:10.1016/j.jad.2006.06.015 Zich JM, Attkisson CC, Greenfield TK (1990) Screening for depression in primary care clinics: the CES-D and the BDI. Int J Psychiatry Med 20(3):259-277. doi:10.2190/LYKR-7VHP-YJEM-MKM2 Peer E, Vosgerau J, Acquisti A (2013) Reputation as a sufficient condition for data quality on Amazon Mechanical Turk. Behav Res Methods 46(4):1023-1031. doi:10.3758/s13428-013-0434-y Litman L, Robinson J, Rosenzweig C (2014) The relationship between motivation, monetary compensation, and data quality among US- and India-based workers on Mechanical Turk. Behav Res Methods 47(2):519-528. doi:10.3758/s13428-014-0483-x Cuijpers P, Boluijt B, van Straten A (2007) Screening of depression in adolescents through the Internet. Eur Child Adolesc Psychiatry 17(1):32-38. doi:10.1007/s00787-007-0631-2 Haringsma R, Engels GI, Beekman ATF, Spinhoven P (2004) The criterion validity of the center for epidemiological studies depression scale (CES-D) in a sample of self-referred elders with depressive symptomatology. Int J Geriatr Psychiatry 19(6):558-563. doi:10.1002/gps.1130 Dodds PS, Harris KD, Kloumann IM, Bliss CA, Danforth CM (2011) Temporal patterns of happiness and information in a global social network: hedonometrics and Twitter. PLoS ONE 6(12):e26752. doi:10.1371/journal.pone.0026752 Reece AG, Reagan AJ, Lix KLM, Dodds PS, Danforth CM, Langer EJ (2016) Forecasting the onset and course of mental illness with Twitter data. arXiv:1608.07740 Detels R (2009) The scope and concerns of public health. Oxford University Press, London Fiske ST, Hauser RM (2014) Protecting human research participants in the age of big data. Proc Natl Acad Sci USA 111(38):13675-13676. doi:10.1073/pnas.1414626111 Lumb D (2016) Scientists release personal data for 70,000 OkCupid profiles. Available at engt.co/2b4NnQ0. Accessed August 7, 2016 Lazer D, Kennedy R, King G, Vespignani A (2014) The parable of Google flu: traps in big data analysis. Science 343(6176):1203-1205. doi:10.1126/science.1248506 Gigerenzer G (2004) Mindless statistics. J Socio-Econ 33(5):587-606. doi:10.1016/j.socec.2004.09.03 Hubbard R, Lindsay RM (2008) Why p-values are not a useful measure of evidence in statistical significance testing. Theory Psychol 18(1):69-88. doi:10.1177/0959354307086923 Morey RD, Hoekstra R, Rouder JN, Lee MD, Wagenmakers EJ (2015) The fallacy of placing confidence in confidence intervals. Psychon Bull Rev 23(1):103-123. doi:10.3758/s13423-015-0947-8 Wasserstein RL, Lazar NA (2016) The ASA’s statement on p-values: context, process, and purpose. Am Stat 70(2):129-133. doi:10.1080/00031305.2016.1154108 Martin A, Quinn K, Park JH (2011) MCMCpack: Markov chain Monte Carlo in R. J Stat Softw 42(9):1-21 Link WA, Eaton MJ (2012) On thinning of chains in MCMC. Methods Ecol Evol 3(1):112-115. doi:10.1111/j.2041-210X.2011.00131.x Christensen R, Johnson W, Branscum A, Hanson TE (2011) Bayesian ideas and data analysis: an introduction for scientists and statisticians. CRC Press, Boca Raton Jeffries H (1961) Theory of probability. Clarendon, Oxford Geweke J (1992) Evaluating the accuracy of sampling-based approaches to the calculation of posterior moments. Vol. 196. Federal Reserve Bank of Minneapolis, Research Department, Minneapolis Gelman A, Rubin DB (1992) Inference from iterative simulation using multiple sequences. Stat Sci 7(4):457-472 Gelman A, Carlin JB, Stern HS, Rubin DB (2014) Bayesian data analysis (Vol. 2). CRC Press, Boca Raton