PT - JOURNAL ARTICLE AU - Stefan Frässle AU - Andre F. Marquand AU - Lianne Schmaal AU - Richard Dinga AU - Dick J. Veltman AU - Nic J.A. van der Wee AU - Marie-José van Tol AU - Dario Schöbi AU - Brenda W.J.H. Penninx AU - Klaas E. Stephan TI - Predicting individual clinical trajectories of depression with generative embedding AID - 10.1101/19006650 DP - 2019 Jan 01 TA - medRxiv PG - 19006650 4099 - http://medrxiv.org/content/early/2019/09/18/19006650.short 4100 - http://medrxiv.org/content/early/2019/09/18/19006650.full AB - Patients with major depressive disorder (MDD) show heterogeneous treatment response and highly variable clinical trajectories: while some patients experience swift and enduring recovery, others show relapsing-remitting or chronic disease course. Predicting individual clinical trajectories at an early disease stage is a key challenge for psychiatry and might facilitate individually tailored interventions. So far, however, reliable predictors at the single-patient level are absent.Here, we evaluated the utility of a machine learning strategy – generative embedding – which combines an interpretable generative model with a discriminative classifier. Specifically, we used functional magnetic resonance imaging (fMRI) data of emotional face perception in 85 MDD patients from the multi-site longitudinal NEtherlands Study of Depression and Anxiety (NESDA) who had been followed up over two years and classified into three subgroups with distinct clinical trajectories. Combining a generative model of effective (directed) connectivity with support vector machines (SVMs), it was possible to predict whether a given patient will experience chronic depression vs. fast remission with a balanced accuracy of 79%. Gradual improvement vs. fast remission could still be predicted above-chance, but less convincingly, with a balanced accuracy of 61%. Importantly, generative embedding outperformed conventional (descriptive) measures such as functional connectivity or local BOLD activity, which did not predict clinical trajectories with above-chance accuracy. Furthermore, the predictive performance of generative embedding could be assigned to a specific network property: the dynamic modulation of connections by the emotional content of the trial-by-trial stimuli. Our findings suggest that a mechanistically informed generative model of a neuronal circuit underlying emotional face perception may have predictive utility for distinguishing disease courses in MDD patients.Competing Interest StatementThe authors have declared no competing interest.Funding StatementThis work was supported by the ETH Zurich Postdoctoral Fellowship Program (SF), the Marie Curie Actions for People COFUND Program (SF), and the University of Zurich Forschungskredit Postdoc (SF), by a NHMRC Career Development Fellowship (1140764; LS), as well as by the René and Susanne Braginsky Foundation (KES) and the University of Zurich (KES). Furthermore, the infrastructure for the NESDA study (www.nesda.nl) is funded through the Geestkracht program of the Netherlands Organisation for Health Research and Development (ZonMw, grant number 10-000-1002) and financial contributions by participating universities and mental health care organizations (VU University Medical Center, GGZ inGeest, Leiden University Medical Center, Leiden University, GGZ Rivierduinen, University Medical Center Groningen, University of Groningen, Lentis, GGZ Friesland, GGZ Drenthe, Rob Giel Onderzoekscentrum).Author DeclarationsAll relevant ethical guidelines have been followed and any necessary IRB and/or ethics committee approvals have been obtained.YesAll necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived.YesAny clinical trials involved have been registered with an ICMJE-approved registry such as ClinicalTrials.gov and the trial ID is included in the manuscript.Not ApplicableI have followed all appropriate research reporting guidelines and uploaded the relevant Equator, ICMJE or other checklist(s) as supplementary files, if applicable.YesThe data utilized in this study are available as part of the NESDA consortium upon reasonable request.