Your browser doesn't support javascript.
loading
An uncertainty-aware, shareable, and transparent neural network architecture for brain-age modeling.
Hahn, Tim; Ernsting, Jan; Winter, Nils R; Holstein, Vincent; Leenings, Ramona; Beisemann, Marie; Fisch, Lukas; Sarink, Kelvin; Emden, Daniel; Opel, Nils; Redlich, Ronny; Repple, Jonathan; Grotegerd, Dominik; Meinert, Susanne; Hirsch, Jochen G; Niendorf, Thoralf; Endemann, Beate; Bamberg, Fabian; Kröncke, Thomas; Bülow, Robin; Völzke, Henry; von Stackelberg, Oyunbileg; Sowade, Ramona Felizitas; Umutlu, Lale; Schmidt, Börge; Caspers, Svenja; Kugel, Harald; Kircher, Tilo; Risse, Benjamin; Gaser, Christian; Cole, James H; Dannlowski, Udo; Berger, Klaus.
Afiliação
  • Hahn T; Institute for Translational Psychiatry, University of Münster, Münster, Germany.
  • Ernsting J; Institute for Translational Psychiatry, University of Münster, Münster, Germany.
  • Winter NR; Faculty of Mathematics and Computer Science, University of Münster, Münster, Germany.
  • Holstein V; Institute for Translational Psychiatry, University of Münster, Münster, Germany.
  • Leenings R; Institute for Translational Psychiatry, University of Münster, Münster, Germany.
  • Beisemann M; Institute for Translational Psychiatry, University of Münster, Münster, Germany.
  • Fisch L; Faculty of Mathematics and Computer Science, University of Münster, Münster, Germany.
  • Sarink K; Department of Statistics, TU Dortmund University, Dortmund, Germany.
  • Emden D; Institute for Translational Psychiatry, University of Münster, Münster, Germany.
  • Opel N; Institute for Translational Psychiatry, University of Münster, Münster, Germany.
  • Redlich R; Institute for Translational Psychiatry, University of Münster, Münster, Germany.
  • Repple J; Institute for Translational Psychiatry, University of Münster, Münster, Germany.
  • Grotegerd D; Interdisciplinary Centre for Clinical Research (IZKF) of the Medical Faculty Münster, University of Münster, Münster, Germany.
  • Meinert S; Institute for Translational Psychiatry, University of Münster, Münster, Germany.
  • Hirsch JG; Department of Psychology, University of Halle, Halle, Germany.
  • Niendorf T; Institute for Translational Psychiatry, University of Münster, Münster, Germany.
  • Endemann B; Institute for Translational Psychiatry, University of Münster, Münster, Germany.
  • Bamberg F; Institute for Translational Psychiatry, University of Münster, Münster, Germany.
  • Kröncke T; Fraunhofer MEVIS, Bremen, Germany.
  • Bülow R; Berlin Ultrahigh Field Facility (B.U.F.F.), NAKO imaging site Berlin, Max-Delbrueck Center for Molecular Medicine in the Helmholtz Association, Berlin, Germany.
  • Völzke H; Berlin Ultrahigh Field Facility (B.U.F.F.), NAKO imaging site Berlin, Max-Delbrueck Center for Molecular Medicine in the Helmholtz Association, Berlin, Germany.
  • von Stackelberg O; Department of Radiology, Medical Center-University of Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, Germany.
  • Sowade RF; Department of Diagnostic and Interventional Radiology, University Hospital Augsburg, Augsburg, Germany.
  • Umutlu L; Institute of Diagnostic Radiology and Neuroradiology, University of Greifswald, Greifswald, Germany.
  • Schmidt B; Institute for Community Medicine, University Medicine Greifswald, Greifswald, Germany.
  • Caspers S; Department of Diagnostic and Interventional Radiology, University Hospital Heidelberg, Heidelberg, Germany.
  • Kugel H; Translational Lung Research Center, Member of the German Lung Research Center, Heidelberg, Germany.
  • Kircher T; Department of Diagnostic and Interventional Radiology, University Hospital Heidelberg, Heidelberg, Germany.
  • Risse B; Translational Lung Research Center, Member of the German Lung Research Center, Heidelberg, Germany.
  • Gaser C; Institute for Medical Informatics, Biometry and Epidemiology, University of Duisburg-Essen, Duisburg, Germany.
  • Cole JH; Institute for Medical Informatics, Biometry and Epidemiology, University of Duisburg-Essen, Duisburg, Germany.
  • Dannlowski U; Institute for Anatomy I, Medical Faculty and University Hospital Düsseldorf, Heinrich Heine University Düsseldorf, 40225 Düsseldorf, Germany.
  • Berger K; Institute of Neuroscience and Medicine (INM-1), Research Centre Jülich, 52425 Jülich, Germany.
Sci Adv ; 8(1): eabg9471, 2022 Jan 07.
Article em En | MEDLINE | ID: mdl-34985964
The deviation between chronological age and age predicted from neuroimaging data has been identified as a sensitive risk marker of cross-disorder brain changes, growing into a cornerstone of biological age research. However, machine learning models underlying the field do not consider uncertainty, thereby confounding results with training data density and variability. Also, existing models are commonly based on homogeneous training sets, often not independently validated, and cannot be shared because of data protection issues. Here, we introduce an uncertainty-aware, shareable, and transparent Monte Carlo dropout composite quantile regression (MCCQR) Neural Network trained on N = 10,691 datasets from the German National Cohort. The MCCQR model provides robust, distribution-free uncertainty quantification in high-dimensional neuroimaging data, achieving lower error rates compared with existing models. In two examples, we demonstrate that it prevents spurious associations and increases power to detect deviant brain aging. We make the pretrained model and code publicly available.

Texto completo: 1 Coleções: 01-internacional Base de dados: MEDLINE Tipo de estudo: Prognostic_studies Idioma: En Ano de publicação: 2022 Tipo de documento: Article

Texto completo: 1 Coleções: 01-internacional Base de dados: MEDLINE Tipo de estudo: Prognostic_studies Idioma: En Ano de publicação: 2022 Tipo de documento: Article