The individual-level, multimodal neural signature of face processing in the fusiform gyrus in autism

345 

Abstract Submission 

Dorothea Floris^1,2, Alberto Llera², Tzvetan Popov¹, Mariam Zabihi³, Carolin Moessnang⁴, Emily Jones⁵, Luke Mason⁵, Rianne Haartsen⁵, Nathalie Holz⁶, Ting Mei², Camille Elleaume⁷, Bruno Hebling Vieira¹, Charlotte Pretzsch⁸, Natalie Forde⁹, Sarah Baumeister⁶, Flavio Dell’Acqua⁸, Sarah Durston¹⁰, Tobias Banaschewski⁶, Christine Ecker¹¹, Rosie Holt¹², Simon Baron-Cohen¹³, Thomas Bourgeron¹⁴, Tony Charman¹⁵, Eva Loth⁸, Declan Murphy⁸, Jan Buitelaar², Christian Beckmann², Nicolas Langer¹, EU AIMS LEAP Group¹⁶

¹Department of Psychology, University of Zurich, Zurich, Switzerland, ²Donders Institute for Brain, Cognition, and Behavior, Radboud University Nijmegen, Nijmegen, NL, Nijmegen, Netherlands, ³MRC Unit Lifelong Health and Aging, University College London, London, UK, London, United Kingdom, ⁴Department of Psychiatry and Psychotherapy, Central Institute of Mental Health, Mannheim, Germany, Mannheim, Germany, ⁵Centre for Brain and Cognitive Development, Birkbeck, University of London, London, UK, London, United Kingdom, ⁶Department of Child and Adolescent Psychiatry, Central Institute of Mental Health, Mannheim, Mannheim, Germany, ⁷Department of Psychology, University of Zurich, Zurich, Zurich, ⁸Department of Forensic and Neurodevelopmental Sciences, IoPPN, London, United Kingdom, ⁹Donders Institute for Brain, Cognition, and Behavior, Radboud University Nijmegen, Nijmegen, NL, Nijmegeb, Netherlands, ¹⁰Department of Psychiatry, Brain Center Rudolf Magnus, University Medical Center Utrecht, Utrecht, Utrecht, Netherlands, ¹¹Department of Child and Adolescent Psychiatry, University Hospital, Goethe University, Frankfurt, Frankfurt, Germany, ¹²Autism Research Centre, Department of Psychiatry, University of Cambridge, Cambridge, UK, Cambridge, United Kingdom, ¹³University of Cambridge, Cambridge, United Kingdom, ¹⁴Institut Pasteur, Human Genetics and Cognitive Functions Unity, Paris, France, Paris, France, ¹⁵Clinical Child Psychology, Department of Psychology, IoPPN, London, United Kingdom, ¹⁶Department of Forensic and Neurodevelopmental Sciences, IoPP, London, United Kingdom

Dorothea Floris  
Department of Psychology, University of Zurich|Donders Institute for Brain, Cognition, and Behavior, Radboud University Nijmegen, Nijmegen, NL
Zurich, Switzerland|Nijmegen, Netherlands

Alberto Llera  
Donders Institute for Brain, Cognition, and Behavior, Radboud University Nijmegen, Nijmegen, NL
Nijmegen, Netherlands

Tzvetan Popov  
Department of Psychology, University of Zurich
Zurich, Switzerland

Mariam Zabihi  
MRC Unit Lifelong Health and Aging, University College London, London, UK
London, United Kingdom

Carolin Moessnang  
Department of Psychiatry and Psychotherapy, Central Institute of Mental Health, Mannheim, Germany
Mannheim, Germany

Emily Jones  
Centre for Brain and Cognitive Development, Birkbeck, University of London, London, UK
London, United Kingdom

Luke Mason  
Centre for Brain and Cognitive Development, Birkbeck, University of London, London, UK
London, United Kingdom

Rianne Haartsen  
Centre for Brain and Cognitive Development, Birkbeck, University of London, London, UK
London, United Kingdom

Nathalie Holz  
Department of Child and Adolescent Psychiatry, Central Institute of Mental Health, Mannheim
Mannheim, Germany

Ting Mei  
Donders Institute for Brain, Cognition, and Behavior, Radboud University Nijmegen, Nijmegen, NL
Nijmegen, Netherlands

Camille Elleaume  
Department of Psychology, University of Zurich
Zurich, Zurich

Bruno Hebling Vieira  
Department of Psychology, University of Zurich
Zurich, Switzerland

Charlotte Pretzsch  
Department of Forensic and Neurodevelopmental Sciences, IoPPN
London, United Kingdom

Natalie Forde  
Donders Institute for Brain, Cognition, and Behavior, Radboud University Nijmegen, Nijmegen, NL
Nijmegeb, Netherlands

Sarah Baumeister  
Department of Child and Adolescent Psychiatry, Central Institute of Mental Health, Mannheim
Mannheim, Germany

Flavio Dell’Acqua  
Department of Forensic and Neurodevelopmental Sciences, IoPPN
London, United Kingdom

Sarah Durston  
Department of Psychiatry, Brain Center Rudolf Magnus, University Medical Center Utrecht, Utrecht
Utrecht, Netherlands

Tobias Banaschewski  
Department of Child and Adolescent Psychiatry, Central Institute of Mental Health, Mannheim
Mannheim, Germany

Christine Ecker  
Department of Child and Adolescent Psychiatry, University Hospital, Goethe University, Frankfurt
Frankfurt, Germany

Rosie Holt  
Autism Research Centre, Department of Psychiatry, University of Cambridge, Cambridge, UK
Cambridge, United Kingdom

Simon Baron-Cohen  
University of Cambridge
Cambridge, United Kingdom

Thomas Bourgeron  
Institut Pasteur, Human Genetics and Cognitive Functions Unity, Paris, France
Paris, France

Tony Charman  
Clinical Child Psychology, Department of Psychology, IoPPN
London, United Kingdom

Eva Loth  
Department of Forensic and Neurodevelopmental Sciences, IoPPN
London, United Kingdom

Declan Murphy  
Department of Forensic and Neurodevelopmental Sciences, IoPPN
London, United Kingdom

Jan Buitelaar  
Donders Institute for Brain, Cognition, and Behavior, Radboud University Nijmegen, Nijmegen, NL
Nijmegen, Netherlands

Christian Beckmann  
Donders Institute for Brain, Cognition, and Behavior, Radboud University Nijmegen, Nijmegen, NL
Nijmegen, Netherlands

Nicolas Langer  
Department of Psychology, University of Zurich
Zurich, Switzerland

EU AIMS LEAP Group  
Department of Forensic and Neurodevelopmental Sciences, IoPP
London, United Kingdom

Face processing is among the most commonly reported social difficulties of autistic individuals (1,2). While its neural underpinnings have been explored extensively across single neuroimaging modalities within key regions of the face processing network, such as the fusiform gyrus (FFG) (3,4), there is still little knowledge about how different structural and functional neurobiological markers are simultaneously implicated in face processing in autism and associated with social functioning. Extracting the joint, shared information across different modalities is essential to better elucidate complex relationships between brain structure and function leading to a more comprehensive understanding of underlying mechanisms of autism.

Here, we leveraged the large multimodal EU-AIMS Longitudinal European Autism Project dataset (5) to study the cross-modal signature of face processing within the FFG across structural magnetic resonance imaging (MRI), resting-state fMRI (rs-fMRI), task-fMRI (based on the Hariri emotional faces task) and electroencephalography (EEG) (recorded when observing facial stimuli) in a sample of 99 autistic and 105 non-autistic individuals between 6-30 years of age. After employing normative modelling (6) using the PCNtoolkit on each imaging modality to derive individual-level deviations from a normative developmental trajectory, unimodal deviation scores were merged using linked independent component (IC) analysis (7). We next tested whether ICs significantly differed between autistic and non-autistic individuals (NAI) using a general linear model and whether multimodal ICs would outperform unimodal ICs in discriminating autistic individuals from NAI using a support vector machine under 10-fold cross-validation. To test the association between multimodal ICs and cognitive, clinical features related to either social or non-social functioning in autism, canonical correlation analysis (CCA) was employed.

In total, 50 independent components were derived, among which one IC showed a significant difference between autistic and non-autistic individuals (t=3.5, pFDR=0.03) (Figure 1). This IC was mostly driven by bilateral rs-fMRI, bilateral structure, right task-fMRI, and left EEG and implicated both face-selective and retinotopic regions of the FFG (Figure 2). Furthermore, comparing areas under the curve with a permutation test, multimodal ICs performed significantly better at differentiating between autistic individuals and NAI (p<0.001). Finally, there was a significant multivariate canonical correlation between multimodal ICs and a set of cognitive, clinical features associated with social function (r=0.65, pFDR=0.008). This was not the case for the association with a set of non-social features.

Results suggest that the FFG is a central region differentially implicated in autistic and non-autistic individuals across a range of imaging modalities and these can simultaneously inform mechanisms associated with core social functioning in autism. These findings further suggest that the discerning signals in this specific brain region are reliably captured through components shared across modalities, emphasizing the multidimensional nature of effects associated with autism. Elucidating a more holistic picture of neural associations of core cognitive and clinical features in autism, will pave the way for the development of more personalised support.

Neurodevelopmental/ Early Life (eg. ADHD, autism) ¹

Multivariate Approaches ²

Autism

FUNCTIONAL MRI

Machine Learning

STRUCTURAL MRI