Generating synthetic task-based brain fingerprints for population neuroscience using deep learning

Task-based functional magnetic resonance imaging (fMRI) reveals individual differences in neural correlates of cognition but faces scalability challenges due to cognitive demands, protocol variability, and limited task coverage in large datasets. Here, we propose DeepTaskGen, a deep-learning approach that synthesizes non-acquired task-based contrast maps from resting-state (rs-) fMRI. We validate this approach using the Human Connectome Project lifespan data, then generate 47 contrast maps from 7 different cognitive tasks for over 20,000 individuals from UK Biobank. DeepTaskGen outperforms several benchmarks in generating synthetic task-contrast maps, achieving superior reconstruction performance while retaining inter-individual variation essential for biomarker development. We further show comparable or superior predictive performance of synthetic maps relative to actual maps and rs-connectomes across diverse demographic, cognitive, and clinical variables. This approach facilitates the study of individual differences and the generation of task-related biomarkers by enabling the generation of arbitrary functional cognitive tasks from readily available rs-fMRI data.

Authors

Institutions

Publication Details

Journal
Communications Biology
Published
2025-11-14
DOI
https://doi.org/10.1038/s42003-025-09158-6
Primary Topic
Functional Brain Connectivity Studies
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Generating synthetic task-based brain fingerprints for population neuroscience using deep learning

Sylvane Desrivières, Bernhard Spanlang, Lars Tjelta Westlye, Verena Kohler et al.
Communications Biology
Functional Brain Connectivity Studies
article

Generating synthetic task-based brain fingerprints for population neuroscience using deep learning

Sylvane Desrivières, Bernhard Spanlang, Lars Tjelta Westlye, Verena Kohler, Kerstin Schepanski, Sören Hese, Andreas Meyer‐Lindenberg, Antoine Bernas, Ulrike H. Taron, Karina Jansone, Hedi Kebir, Nicholas Clinton, Ragnhild Eek Brandlistuen, Myrto Patraskaki, Mira Tschorn, Esther Hitchen, Sara Fernández, Jiacan Yuan, Tobias Banaschewski, Jaime Gallego, Carina M. Mathey, Emanuel Schwarz, Ole A. Andreassen, Rikka Kjelkenes, Nina Christmann, Per Hoffmann, Heike Tost, Helga Ask, Karen Schmitt, Andreas Heinz, Emin Serin, Tamara Schikowski, Sven Twardziok, Francisco José Eiroá‐Orosa, Paul Renner, Nathalie E. Holz, Bernd Carsten Stahl, Jean-Charles Roy, Roland Eils, Viktor Jirsa, Sarah Jane Böttger, Michael A. Rapp, Kofoworola Agunbiade, Gaia Novarino, Rieke Aden, Henrik Walter, André F. Marquand, Spase Petkoski, Stefanie Heilmann‐Heimbach, Peter Sommer, Tianye Jia, Dennis van der Meer, Allan H. Young, Xiao Chang, Tristram A. Lett, Kerstin Ritter, Markus Ralser, Mel Slater, Johannes Wilbertz, Frauke Nees, Guillem Feixas, Reiya Itatani, George Ogoh, Elli Polemiti, Isabelle Claus, Argyris Stringaris, Zuo Zhang, Markus M. Nöthen, Maja Neidhart, Charlie Pearmund, Di Chen, Yanqing Zhang, Yuzhu Li, Xinyang Yu, Günter Schumann, Paul Thompson, Nilakshi Vaidya, Beke Seefried, Andreas J. Forstner, Anastasios-Polykarpos Athanasiadis, Marcel Jentsch, Ameli Schwalber, Yunman Xia, Tatjana Schütz, Álvaro Pastor, Vince D. Calhoun, Yuxiang Dai
article en

Abstract

Task-based functional magnetic resonance imaging (fMRI) reveals individual differences in neural correlates of cognition but faces scalability challenges due to cognitive demands, protocol variability, and limited task coverage in large datasets. Here, we propose DeepTaskGen, a deep-learning approach that synthesizes non-acquired task-based contrast maps from resting-state (rs-) fMRI. We validate this approach using the Human Connectome Project lifespan data, then generate 47 contrast maps from 7 different cognitive tasks for over 20,000 individuals from UK Biobank. DeepTaskGen outperforms several benchmarks in generating synthetic task-contrast maps, achieving superior reconstruction performance while retaining inter-individual variation essential for biomarker development. We further show comparable or superior predictive performance of synthetic maps relative to actual maps and rs-connectomes across diverse demographic, cognitive, and clinical variables. This approach facilitates the study of individual differences and the generation of task-related biomarkers by enabling the generation of arbitrary functional cognitive tasks from readily available rs-fMRI data.

Communications BiologyVol. 8(1)
University of Southern California (US), Oslo University Hospital (NO), Google (United States) (US), Georgia Institute of Technology (US), Norwegian Institute of Public Health (NO), University of Bonn (DE), University of Nottingham (GB), Radboud University Nijmegen (NL), Emory University (US), Institute of Science and Technology Austria (AT), University of Potsdam (DE), Georgia State University (US), King's College London (GB), University of Oslo (NO), South London and Maudsley NHS Foundation Trust (GB), Aix-Marseille Université (FR), Heidelberg University (DE), Fudan University (CN), University Hospital Bonn (DE), University Hospital Heidelberg (DE), Humboldt-Universität zu Berlin (DE), Leibniz Institute of Environmental Medicine (DE), Central Institute of Mental Health (DE), University Hospital Schleswig-Holstein (DE), Center for Translational Research in Neuroimaging and Data Science (US), Berlin Institute of Health at Charité - Universitätsmedizin Berlin (DE), Hertie Institute for Clinical Brain Research (DE), Concentris (Germany) (DE), Bernstein Center for Computational Neuroscience Berlin (DE), Shanghai Center for Brain Science and Brain-Inspired Technology (CN), Universitat de Barcelona (ES), Freie Universität Berlin (DE), Friedrich Schiller University Jena (DE), University of Tübingen (DE), University of Lübeck (DE)
Government of the United Kingdom, UK Research and Innovation, European Commission, Deutsche Forschungsgemeinschaft, National Institutes of Health, McDonnell Center for Systems Neuroscience, HORIZON EUROPE Framework Programme
Openalex Percentile: Top 66%
Functional Brain Connectivity Studies
DOIOpen Access PDFOpenAlex
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.