Normative modelling using deep autoencoders: a multi-cohort study on mild cognitive impairment and Alzheimer’s disease
Authors:
- Walter H. L. PinayaWPWalter H. L. Pinaya
- Cristina ScarpazzaCSCristina Scarpazza
- Rafael Garcia-DiasRGRafael Garcia-Dias
- Sandra VieiraSVSandra Vieira
- Lea BaeckerLBLea Baecker
- Pedro F. da CostaPCPedro F. da Costa
- Alberto RedolfiARAlberto Redolfi
- Giovanni B. FrisoniGFGiovanni B. Frisoni
- Michela PievaniMPMichela Pievani
- Vince D. CalhounVCVince D. Calhoun
- João R. SatoJSJoão R. Sato
- Andrea MechelliAMAndrea Mechelli
Preprint
Abstract
Normative modelling is an emerging method for quantifying how individuals deviate from the healthy populational pattern. Several machine learning models have been implemented to develop normative models to investigate brain disorders, including regression, support vector machines and Gaussian process models. With the advance of deep learning technology, the use of deep neural networks has also been proposed. In this study, we assessed normative models based on deep autoencoders using structural neuroimaging data from patients with Alzheimer’s disease (n=206) and mild cognitive impairment (n=354). We first trained the autoencoder on an independent dataset (UK Biobank dataset) with 11,034 healthy controls. Then, we estimated how each patient deviated from this norm and established which brain regions were associated to this deviation. Finally, we compared the performance of our normative model against traditional classifiers. As expected, we found that patients exhibited deviations according to the severity of their clinical condition. The model identified medial temporal regions, including the hippocampus, and the ventricular system as critical regions for the calculation of the deviation score. Overall, the normative model had comparable cross-cohort generalizability to traditional classifiers. In order to promote open science, we are making all scripts and the trained models available to the wider research community.Keywords
This publication has 44 references indexed in Scilit:
- Deep neural networks2020
- Conceptualizing mental disorders as deviations from normative functioningMolecular Psychiatry, 2019
- Mapping the Heterogeneous Phenotype of Schizophrenia and Bipolar Disorder Using Normative ModelsJAMA Psychiatry, 2018
- Using deep autoencoders to identify abnormal brain structural patterns in neuropsychiatric disorders: A large‐scale multi‐sample studyHuman Brain Mapping, 2018
- A spatio-temporal reference model of the aging brainNeuroImage, 2018
- Understanding Heterogeneity in Clinical Cohorts Using Normative Models: Beyond Case-Control StudiesBiological Psychiatry, 2016
- Deep learningNature, 2015
- Individualized Gaussian process-based prediction and detection of local and global gray matter abnormalities in elderly subjectsNeuroImage, 2014
- Fully-Automated Volumetric MRI with Normative Ranges: Translation to Clinical PracticeBehavioural Neurology, 2009
- Voxel-based morphometry in Alzheimer’s diseaseExpert Review of Neurotherapeutics, 2008