Cargando…
Multimodal Machine Learning Workflows for Prediction of Psychosis in Patients With Clinical High-Risk Syndromes and Recent-Onset Depression
IMPORTANCE: Diverse models have been developed to predict psychosis in patients with clinical high-risk (CHR) states. Whether prediction can be improved by efficiently combining clinical and biological models and by broadening the risk spectrum to young patients with depressive syndromes remains unc...
Autores principales: | , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , |
---|---|
Formato: | Online Artículo Texto |
Lenguaje: | English |
Publicado: |
American Medical Association
2020
|
Materias: | |
Acceso en línea: | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7711566/ https://www.ncbi.nlm.nih.gov/pubmed/33263726 http://dx.doi.org/10.1001/jamapsychiatry.2020.3604 |
_version_ | 1783618175034720256 |
---|---|
author | Koutsouleris, Nikolaos Dwyer, Dominic B. Degenhardt, Franziska Maj, Carlo Urquijo-Castro, Maria Fernanda Sanfelici, Rachele Popovic, David Oeztuerk, Oemer Haas, Shalaila S. Weiske, Johanna Ruef, Anne Kambeitz-Ilankovic, Lana Antonucci, Linda A. Neufang, Susanne Schmidt-Kraepelin, Christian Ruhrmann, Stephan Penzel, Nora Kambeitz, Joseph Haidl, Theresa K. Rosen, Marlene Chisholm, Katharine Riecher-Rössler, Anita Egloff, Laura Schmidt, André Andreou, Christina Hietala, Jarmo Schirmer, Timo Romer, Georg Walger, Petra Franscini, Maurizia Traber-Walker, Nina Schimmelmann, Benno G. Flückiger, Rahel Michel, Chantal Rössler, Wulf Borisov, Oleg Krawitz, Peter M. Heekeren, Karsten Buechler, Roman Pantelis, Christos Falkai, Peter Salokangas, Raimo K. R. Lencer, Rebekka Bertolino, Alessandro Borgwardt, Stefan Noethen, Markus Brambilla, Paolo Wood, Stephen J. Upthegrove, Rachel Schultze-Lutter, Frauke Theodoridou, Anastasia Meisenzahl, Eva |
author_facet | Koutsouleris, Nikolaos Dwyer, Dominic B. Degenhardt, Franziska Maj, Carlo Urquijo-Castro, Maria Fernanda Sanfelici, Rachele Popovic, David Oeztuerk, Oemer Haas, Shalaila S. Weiske, Johanna Ruef, Anne Kambeitz-Ilankovic, Lana Antonucci, Linda A. Neufang, Susanne Schmidt-Kraepelin, Christian Ruhrmann, Stephan Penzel, Nora Kambeitz, Joseph Haidl, Theresa K. Rosen, Marlene Chisholm, Katharine Riecher-Rössler, Anita Egloff, Laura Schmidt, André Andreou, Christina Hietala, Jarmo Schirmer, Timo Romer, Georg Walger, Petra Franscini, Maurizia Traber-Walker, Nina Schimmelmann, Benno G. Flückiger, Rahel Michel, Chantal Rössler, Wulf Borisov, Oleg Krawitz, Peter M. Heekeren, Karsten Buechler, Roman Pantelis, Christos Falkai, Peter Salokangas, Raimo K. R. Lencer, Rebekka Bertolino, Alessandro Borgwardt, Stefan Noethen, Markus Brambilla, Paolo Wood, Stephen J. Upthegrove, Rachel Schultze-Lutter, Frauke Theodoridou, Anastasia Meisenzahl, Eva |
author_sort | Koutsouleris, Nikolaos |
collection | PubMed |
description | IMPORTANCE: Diverse models have been developed to predict psychosis in patients with clinical high-risk (CHR) states. Whether prediction can be improved by efficiently combining clinical and biological models and by broadening the risk spectrum to young patients with depressive syndromes remains unclear. OBJECTIVES: To evaluate whether psychosis transition can be predicted in patients with CHR or recent-onset depression (ROD) using multimodal machine learning that optimally integrates clinical and neurocognitive data, structural magnetic resonance imaging (sMRI), and polygenic risk scores (PRS) for schizophrenia; to assess models’ geographic generalizability; to test and integrate clinicians’ predictions; and to maximize clinical utility by building a sequential prognostic system. DESIGN, SETTING, AND PARTICIPANTS: This multisite, longitudinal prognostic study performed in 7 academic early recognition services in 5 European countries followed up patients with CHR syndromes or ROD and healthy volunteers. The referred sample of 167 patients with CHR syndromes and 167 with ROD was recruited from February 1, 2014, to May 31, 2017, of whom 26 (23 with CHR syndromes and 3 with ROD) developed psychosis. Patients with 18-month follow-up (n = 246) were used for model training and leave-one-site-out cross-validation. The remaining 88 patients with nontransition served as the validation of model specificity. Three hundred thirty-four healthy volunteers provided a normative sample for prognostic signature evaluation. Three independent Swiss projects contributed a further 45 cases with psychosis transition and 600 with nontransition for the external validation of clinical-neurocognitive, sMRI-based, and combined models. Data were analyzed from January 1, 2019, to March 31, 2020. MAIN OUTCOMES AND MEASURES: Accuracy and generalizability of prognostic systems. RESULTS: A total of 668 individuals (334 patients and 334 controls) were included in the analysis (mean [SD] age, 25.1 [5.8] years; 354 [53.0%] female and 314 [47.0%] male). Clinicians attained a balanced accuracy of 73.2% by effectively ruling out (specificity, 84.9%) but ineffectively ruling in (sensitivity, 61.5%) psychosis transition. In contrast, algorithms showed high sensitivity (76.0%-88.0%) but low specificity (53.5%-66.8%). A cybernetic risk calculator combining all algorithmic and human components predicted psychosis with a balanced accuracy of 85.5% (sensitivity, 84.6%; specificity, 86.4%). In comparison, an optimal prognostic workflow produced a balanced accuracy of 85.9% (sensitivity, 84.6%; specificity, 87.3%) at a much lower diagnostic burden by sequentially integrating clinical-neurocognitive, expert-based, PRS-based, and sMRI-based risk estimates as needed for the given patient. Findings were supported by good external validation results. CONCLUSIONS AND RELEVANCE: These findings suggest that psychosis transition can be predicted in a broader risk spectrum by sequentially integrating algorithms’ and clinicians’ risk estimates. For clinical translation, the proposed workflow should undergo large-scale international validation. |
format | Online Article Text |
id | pubmed-7711566 |
institution | National Center for Biotechnology Information |
language | English |
publishDate | 2020 |
publisher | American Medical Association |
record_format | MEDLINE/PubMed |
spelling | pubmed-77115662020-12-03 Multimodal Machine Learning Workflows for Prediction of Psychosis in Patients With Clinical High-Risk Syndromes and Recent-Onset Depression Koutsouleris, Nikolaos Dwyer, Dominic B. Degenhardt, Franziska Maj, Carlo Urquijo-Castro, Maria Fernanda Sanfelici, Rachele Popovic, David Oeztuerk, Oemer Haas, Shalaila S. Weiske, Johanna Ruef, Anne Kambeitz-Ilankovic, Lana Antonucci, Linda A. Neufang, Susanne Schmidt-Kraepelin, Christian Ruhrmann, Stephan Penzel, Nora Kambeitz, Joseph Haidl, Theresa K. Rosen, Marlene Chisholm, Katharine Riecher-Rössler, Anita Egloff, Laura Schmidt, André Andreou, Christina Hietala, Jarmo Schirmer, Timo Romer, Georg Walger, Petra Franscini, Maurizia Traber-Walker, Nina Schimmelmann, Benno G. Flückiger, Rahel Michel, Chantal Rössler, Wulf Borisov, Oleg Krawitz, Peter M. Heekeren, Karsten Buechler, Roman Pantelis, Christos Falkai, Peter Salokangas, Raimo K. R. Lencer, Rebekka Bertolino, Alessandro Borgwardt, Stefan Noethen, Markus Brambilla, Paolo Wood, Stephen J. Upthegrove, Rachel Schultze-Lutter, Frauke Theodoridou, Anastasia Meisenzahl, Eva JAMA Psychiatry Original Investigation IMPORTANCE: Diverse models have been developed to predict psychosis in patients with clinical high-risk (CHR) states. Whether prediction can be improved by efficiently combining clinical and biological models and by broadening the risk spectrum to young patients with depressive syndromes remains unclear. OBJECTIVES: To evaluate whether psychosis transition can be predicted in patients with CHR or recent-onset depression (ROD) using multimodal machine learning that optimally integrates clinical and neurocognitive data, structural magnetic resonance imaging (sMRI), and polygenic risk scores (PRS) for schizophrenia; to assess models’ geographic generalizability; to test and integrate clinicians’ predictions; and to maximize clinical utility by building a sequential prognostic system. DESIGN, SETTING, AND PARTICIPANTS: This multisite, longitudinal prognostic study performed in 7 academic early recognition services in 5 European countries followed up patients with CHR syndromes or ROD and healthy volunteers. The referred sample of 167 patients with CHR syndromes and 167 with ROD was recruited from February 1, 2014, to May 31, 2017, of whom 26 (23 with CHR syndromes and 3 with ROD) developed psychosis. Patients with 18-month follow-up (n = 246) were used for model training and leave-one-site-out cross-validation. The remaining 88 patients with nontransition served as the validation of model specificity. Three hundred thirty-four healthy volunteers provided a normative sample for prognostic signature evaluation. Three independent Swiss projects contributed a further 45 cases with psychosis transition and 600 with nontransition for the external validation of clinical-neurocognitive, sMRI-based, and combined models. Data were analyzed from January 1, 2019, to March 31, 2020. MAIN OUTCOMES AND MEASURES: Accuracy and generalizability of prognostic systems. RESULTS: A total of 668 individuals (334 patients and 334 controls) were included in the analysis (mean [SD] age, 25.1 [5.8] years; 354 [53.0%] female and 314 [47.0%] male). Clinicians attained a balanced accuracy of 73.2% by effectively ruling out (specificity, 84.9%) but ineffectively ruling in (sensitivity, 61.5%) psychosis transition. In contrast, algorithms showed high sensitivity (76.0%-88.0%) but low specificity (53.5%-66.8%). A cybernetic risk calculator combining all algorithmic and human components predicted psychosis with a balanced accuracy of 85.5% (sensitivity, 84.6%; specificity, 86.4%). In comparison, an optimal prognostic workflow produced a balanced accuracy of 85.9% (sensitivity, 84.6%; specificity, 87.3%) at a much lower diagnostic burden by sequentially integrating clinical-neurocognitive, expert-based, PRS-based, and sMRI-based risk estimates as needed for the given patient. Findings were supported by good external validation results. CONCLUSIONS AND RELEVANCE: These findings suggest that psychosis transition can be predicted in a broader risk spectrum by sequentially integrating algorithms’ and clinicians’ risk estimates. For clinical translation, the proposed workflow should undergo large-scale international validation. American Medical Association 2020-12-02 2021-02 /pmc/articles/PMC7711566/ /pubmed/33263726 http://dx.doi.org/10.1001/jamapsychiatry.2020.3604 Text en Copyright 2020 Koutsouleris N et al. JAMA Psychiatry. http://creativecommons.org/licenses/by/4.0/ This is an open access article distributed under the terms of the CC-BY License. |
spellingShingle | Original Investigation Koutsouleris, Nikolaos Dwyer, Dominic B. Degenhardt, Franziska Maj, Carlo Urquijo-Castro, Maria Fernanda Sanfelici, Rachele Popovic, David Oeztuerk, Oemer Haas, Shalaila S. Weiske, Johanna Ruef, Anne Kambeitz-Ilankovic, Lana Antonucci, Linda A. Neufang, Susanne Schmidt-Kraepelin, Christian Ruhrmann, Stephan Penzel, Nora Kambeitz, Joseph Haidl, Theresa K. Rosen, Marlene Chisholm, Katharine Riecher-Rössler, Anita Egloff, Laura Schmidt, André Andreou, Christina Hietala, Jarmo Schirmer, Timo Romer, Georg Walger, Petra Franscini, Maurizia Traber-Walker, Nina Schimmelmann, Benno G. Flückiger, Rahel Michel, Chantal Rössler, Wulf Borisov, Oleg Krawitz, Peter M. Heekeren, Karsten Buechler, Roman Pantelis, Christos Falkai, Peter Salokangas, Raimo K. R. Lencer, Rebekka Bertolino, Alessandro Borgwardt, Stefan Noethen, Markus Brambilla, Paolo Wood, Stephen J. Upthegrove, Rachel Schultze-Lutter, Frauke Theodoridou, Anastasia Meisenzahl, Eva Multimodal Machine Learning Workflows for Prediction of Psychosis in Patients With Clinical High-Risk Syndromes and Recent-Onset Depression |
title | Multimodal Machine Learning Workflows for Prediction of Psychosis in Patients With Clinical High-Risk Syndromes and Recent-Onset Depression |
title_full | Multimodal Machine Learning Workflows for Prediction of Psychosis in Patients With Clinical High-Risk Syndromes and Recent-Onset Depression |
title_fullStr | Multimodal Machine Learning Workflows for Prediction of Psychosis in Patients With Clinical High-Risk Syndromes and Recent-Onset Depression |
title_full_unstemmed | Multimodal Machine Learning Workflows for Prediction of Psychosis in Patients With Clinical High-Risk Syndromes and Recent-Onset Depression |
title_short | Multimodal Machine Learning Workflows for Prediction of Psychosis in Patients With Clinical High-Risk Syndromes and Recent-Onset Depression |
title_sort | multimodal machine learning workflows for prediction of psychosis in patients with clinical high-risk syndromes and recent-onset depression |
topic | Original Investigation |
url | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7711566/ https://www.ncbi.nlm.nih.gov/pubmed/33263726 http://dx.doi.org/10.1001/jamapsychiatry.2020.3604 |
work_keys_str_mv | AT koutsoulerisnikolaos multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT dwyerdominicb multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT degenhardtfranziska multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT majcarlo multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT urquijocastromariafernanda multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT sanfelicirachele multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT popovicdavid multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT oeztuerkoemer multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT haasshalailas multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT weiskejohanna multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT ruefanne multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT kambeitzilankoviclana multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT antonuccilindaa multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT neufangsusanne multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT schmidtkraepelinchristian multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT ruhrmannstephan multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT penzelnora multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT kambeitzjoseph multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT haidltheresak multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT rosenmarlene multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT chisholmkatharine multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT riecherrossleranita multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT eglofflaura multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT schmidtandre multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT andreouchristina multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT hietalajarmo multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT schirmertimo multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT romergeorg multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT walgerpetra multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT franscinimaurizia multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT traberwalkernina multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT schimmelmannbennog multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT fluckigerrahel multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT michelchantal multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT rosslerwulf multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT borisovoleg multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT krawitzpeterm multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT heekerenkarsten multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT buechlerroman multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT pantelischristos multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT falkaipeter multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT salokangasraimokr multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT lencerrebekka multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT bertolinoalessandro multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT borgwardtstefan multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT noethenmarkus multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT brambillapaolo multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT woodstephenj multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT upthegroverachel multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT schultzelutterfrauke multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT theodoridouanastasia multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression AT meisenzahleva multimodalmachinelearningworkflowsforpredictionofpsychosisinpatientswithclinicalhighrisksyndromesandrecentonsetdepression |