Uncovering the heterogeneity and temporal complexity of neurodegenerative diseases with Subtype and Stage Inference


Alexandra L. Young, University College London
Razvan V. Marinescu, University College London
Neil P. Oxtoby, University College London
Martina Bocchetta, UCL Queen Square Institute of Neurology
Keir Yong, UCL Queen Square Institute of Neurology
Nicholas C. Firth, University College London
David M. Cash, University College London
David L. Thomas, UCL Queen Square Institute of Neurology
Katrina M. Dick, UCL Queen Square Institute of Neurology
Jorge Cardoso, University College London
John van Swieten, Erasmus MC
Barbara Borroni, Università degli Studi di Brescia
Daniela Galimberti, Università degli Studi di Milano
Mario Masellis, University of Toronto
Maria Carmela Tartaglia, University of Toronto
James B. Rowe, University of Cambridge
Caroline Graff, Karolinska Institutet
Fabrizio Tagliavini, Foundation IRCCS Neurological Institute "C. Besta"
Giovanni B. Frisoni, Université de Genève
Robert Laforce, Université Laval
Elizabeth Finger, The University of Western Ontario
Alexandre de Mendonça, Faculdade de Medicina, Universidade de Lisboa
Sandro Sorbi, Università degli Studi di Firenze
Jason D. Warren, UCL Queen Square Institute of Neurology
Sebastian Crutch, UCL Queen Square Institute of Neurology
Nick C. Fox, UCL Queen Square Institute of Neurology
Sebastien Ourselin, University College London
Jonathan M. Schott, UCL Queen Square Institute of Neurology
Jonathan D. Rohrer, UCL Queen Square Institute of Neurology
Daniel C. Alexander, University College London
Christin Andersson, Karolinska Institutet
Silvana Archetti, Civic Hospital of Brescia

Document Type


Publication Date



Nature Communications





URL with Digital Object Identifier



The heterogeneity of neurodegenerative diseases is a key confound to disease understanding and treatment development, as study cohorts typically include multiple phenotypes on distinct disease trajectories. Here we introduce a machine-learning technique—Subtype and Stage Inference (SuStaIn)—able to uncover data-driven disease phenotypes with distinct temporal progression patterns, from widely available cross-sectional patient studies. Results from imaging studies in two neurodegenerative diseases reveal subgroups and their distinct trajectories of regional neurodegeneration. In genetic frontotemporal dementia, SuStaIn identifies genotypes from imaging alone, validating its ability to identify subtypes; further the technique reveals within-genotype heterogeneity. In Alzheimer’s disease, SuStaIn uncovers three subtypes, uniquely characterising their temporal complexity. SuStaIn provides fine-grained patient stratification, which substantially enhances the ability to predict conversion between diagnostic categories over standard models that ignore subtype (p = 7.18 × 10−4) or temporal stage (p = 3.96 × 10−5). SuStaIn offers new promise for enabling disease subtype discovery and precision medicine.

This document is currently not available here.