Utilizing Machine Learning on Internet Search Activity to Support the Diagnostic Process and Relapse Detection in Young Individuals With Early Psychosis: Feasibility Study.
Google
diagnostic prediction
digital biomarkers
digital data
digital phenotyping
internet search activity
machine learning
relapse prediction
schizophrenia spectrum disorders
Journal
JMIR mental health
ISSN: 2368-7959
Titre abrégé: JMIR Ment Health
Pays: Canada
ID NLM: 101658926
Informations de publication
Date de publication:
01 Sep 2020
01 Sep 2020
Historique:
received:
15
04
2020
accepted:
23
07
2020
revised:
20
07
2020
entrez:
2
9
2020
pubmed:
2
9
2020
medline:
2
9
2020
Statut:
epublish
Résumé
Psychiatry is nearly entirely reliant on patient self-reporting, and there are few objective and reliable tests or sources of collateral information available to help diagnostic and assessment procedures. Technology offers opportunities to collect objective digital data to complement patient experience and facilitate more informed treatment decisions. We aimed to develop computational algorithms based on internet search activity designed to support diagnostic procedures and relapse identification in individuals with schizophrenia spectrum disorders. We extracted 32,733 time-stamped search queries across 42 participants with schizophrenia spectrum disorders and 74 healthy volunteers between the ages of 15 and 35 (mean 24.4 years, 44.0% male), and built machine-learning diagnostic and relapse classifiers utilizing the timing, frequency, and content of online search activity. Classifiers predicted a diagnosis of schizophrenia spectrum disorders with an area under the curve value of 0.74 and predicted a psychotic relapse in individuals with schizophrenia spectrum disorders with an area under the curve of 0.71. Compared with healthy participants, those with schizophrenia spectrum disorders made fewer searches and their searches consisted of fewer words. Prior to a relapse hospitalization, participants with schizophrenia spectrum disorders were more likely to use words related to hearing, perception, and anger, and were less likely to use words related to health. Online search activity holds promise for gathering objective and easily accessed indicators of psychiatric symptoms. Utilizing search activity as collateral behavioral health information would represent a major advancement in efforts to capitalize on objective digital data to improve mental health monitoring.
Sections du résumé
BACKGROUND
BACKGROUND
Psychiatry is nearly entirely reliant on patient self-reporting, and there are few objective and reliable tests or sources of collateral information available to help diagnostic and assessment procedures. Technology offers opportunities to collect objective digital data to complement patient experience and facilitate more informed treatment decisions.
OBJECTIVE
OBJECTIVE
We aimed to develop computational algorithms based on internet search activity designed to support diagnostic procedures and relapse identification in individuals with schizophrenia spectrum disorders.
METHODS
METHODS
We extracted 32,733 time-stamped search queries across 42 participants with schizophrenia spectrum disorders and 74 healthy volunteers between the ages of 15 and 35 (mean 24.4 years, 44.0% male), and built machine-learning diagnostic and relapse classifiers utilizing the timing, frequency, and content of online search activity.
RESULTS
RESULTS
Classifiers predicted a diagnosis of schizophrenia spectrum disorders with an area under the curve value of 0.74 and predicted a psychotic relapse in individuals with schizophrenia spectrum disorders with an area under the curve of 0.71. Compared with healthy participants, those with schizophrenia spectrum disorders made fewer searches and their searches consisted of fewer words. Prior to a relapse hospitalization, participants with schizophrenia spectrum disorders were more likely to use words related to hearing, perception, and anger, and were less likely to use words related to health.
CONCLUSIONS
CONCLUSIONS
Online search activity holds promise for gathering objective and easily accessed indicators of psychiatric symptoms. Utilizing search activity as collateral behavioral health information would represent a major advancement in efforts to capitalize on objective digital data to improve mental health monitoring.
Identifiants
pubmed: 32870161
pii: v7i9e19348
doi: 10.2196/19348
pmc: PMC7492982
doi:
Types de publication
Journal Article
Langues
eng
Pagination
e19348Subventions
Organisme : NIMH NIH HHS
ID : K23 MH120505
Pays : United States
Organisme : NIGMS NIH HHS
ID : R01 GM112697
Pays : United States
Organisme : NIMH NIH HHS
ID : R01 MH117172
Pays : United States
Informations de copyright
©Michael Leo Birnbaum, Prathamesh "Param" Kulkarni, Anna Van Meter, Victor Chen, Asra F Rizvi, Elizabeth Arenare, Munmun De Choudhury, John M Kane. Originally published in JMIR Mental Health (http://mental.jmir.org), 01.09.2020.
Références
J Med Internet Res. 2017 Aug 14;19(8):e289
pubmed: 28807891
NPJ Schizophr. 2019 Oct 7;5(1):17
pubmed: 31591400
Schizophr Res. 2012 Aug;139(1-3):116-28
pubmed: 22658527
Sci Rep. 2017 Oct 11;7(1):13006
pubmed: 29021528
Psychiatry Res. 2018 Jan;259:482-487
pubmed: 29154169
Arch Gen Psychiatry. 1999 Mar;56(3):241-7
pubmed: 10078501
J Oncol Pract. 2016 Aug;12(8):737-44
pubmed: 27271506
Adv Exp Med Biol. 2019;1118:135-162
pubmed: 30747421
Psychiatr Serv. 2015 Jul;66(7):753-6
pubmed: 25588418
Am J Psychiatry. 1980 Jul;137(7):801-5
pubmed: 6104444
Schizophr Res. 2018 Mar;193:3-10
pubmed: 28634088
Psychiatr Serv. 2018 Dec 1;69(12):1259-1263
pubmed: 30256181
J Clin Psychiatry. 2010 Sep;71(9):1115-24
pubmed: 20923620
Biomed Inform Insights. 2018 Aug 27;10:1178222618792860
pubmed: 30158822
J Med Internet Res. 2018 Jul 20;20(7):e241
pubmed: 30030209
Proc SIGCHI Conf Hum Factor Comput Syst. 2016 May;2016:2098-2110
pubmed: 29082385
Psychiatry Res. 2017 Jul;253:240-248
pubmed: 28395229
Schizophr Res. 2016 Jul;174(1-3):165-171
pubmed: 27131912
Psychol Med. 2010 Oct;40(10):1585-97
pubmed: 20236571
Neuropsychol Rev. 2018 Dec;28(4):509-533
pubmed: 30343458
Med J Aust. 2010 Jun 7;192(S11):S22-6
pubmed: 20528703
NPJ Digit Med. 2018 Apr 23;1:8
pubmed: 31304293
NPJ Digit Med. 2018 Aug 22;1:37
pubmed: 31304319
Early Interv Psychiatry. 2017 Aug;11(4):290-295
pubmed: 25808317
JAMA Psychiatry. 2018 Jun 1;75(6):555-565
pubmed: 29800949
Cochrane Database Syst Rev. 2015 Jan 25;1:CD010653
pubmed: 25879096
Science. 2011 Sep 30;333(6051):1878-81
pubmed: 21960633
Brain Neurosci Adv. 2017 Jan 1;1:2398212817744501
pubmed: 29270466
Expert Rev Proteomics. 2017 Sep;14(9):809-824
pubmed: 28870126
Schizophr Res. 2005 Dec 1;80(1):107-11
pubmed: 16125373
Biol Psychiatry. 2001 Dec 1;50(11):884-97
pubmed: 11743943
Popul Health Metr. 2011 Aug 04;9:29
pubmed: 21816105
Neurosci Lett. 2018 Mar 16;669:59-67
pubmed: 27717830
JAMA Oncol. 2017 Mar 1;3(3):398-401
pubmed: 27832243
J Affect Disord. 2019 Jun 1;252:130-134
pubmed: 30981056
Soc Sci Med. 2005 Oct;61(8):1821-7
pubmed: 16029778
Metabolism. 2015 Mar;64(3 Suppl 1):S11-5
pubmed: 25467847
Stud Health Technol Inform. 2007;129(Pt 2):1112-6
pubmed: 17911888
Schizophr Res. 2019 Jun;208:167-172
pubmed: 30940400
Jpn J Psychiatry Neurol. 1993 Dec;47(4):753-75
pubmed: 7911165
Artif Intell Med. 2019 Aug;99:101704
pubmed: 31606109
Schizophr Res. 2019 Jun;208:105-113
pubmed: 30979665
J Med Internet Res. 2010 Dec 19;12(5):e70
pubmed: 21169176
Psychol Med. 1989 Aug;19(3):649-56
pubmed: 2798634
Isr J Psychiatry Relat Sci. 2014;51(1):54-62
pubmed: 24858635
Arch Gen Psychiatry. 2001 Aug;58(8):787-94
pubmed: 11483146
BMC Psychiatry. 2010 Jan 07;10:2
pubmed: 20059765
Proc Natl Acad Sci U S A. 2018 Oct 30;115(44):11203-11208
pubmed: 30322910
Biol Psychiatry Cogn Neurosci Neuroimaging. 2018 Mar;3(3):223-230
pubmed: 29486863
NPJ Schizophr. 2015 Mar 04;1:14005
pubmed: 27336027
Psychiatr Rehabil J. 2015 Sep;38(3):218-226
pubmed: 25844912