A deep learning based approach identifies regions more relevant than resting‐state networks to the prediction of general intelligence from resting‐state fMRI. Issue 18 (29th September 2021)

• Record Type: Journal Article
• Authors: Hebling Vieira, Bruno; Dubois, Julien; Calhoun, Vince D.; Garrido Salmon, Carlos Ernesto
• Is Part Of: Human brain mapping. Volume 42:Issue 18(2021)
• Page Start: 5873
• Page End: 5887
• Publication Date: 2021-09-29
• Subjects: brain‐behavior -- deep learning -- fMRI -- intelligence -- resting‐state; Brain mapping -- Periodicals; 611.81
• DOI: 10.1002/hbm.25656
• Languages: English
• ISSNs: 1065-9471

Abstract

Abstract: Prediction of cognitive ability latent factors such as general intelligence from neuroimaging has elucidated questions pertaining to their neural origins. However, predicting general intelligence from functional connectivity limit hypotheses to that specific domain, being agnostic to time‐distributed features and dynamics. We used an ensemble of recurrent neural networks to circumvent this limitation, bypassing feature extraction, to predict general intelligence from resting‐state functional magnetic resonance imaging regional signals of a large sample ( n = 873) of Human Connectome Project adult subjects. Ablating common resting‐state networks (RSNs) and measuring degradation in performance, we show that model reliance can be mostly explained by network size. Using our approach based on the temporal variance of saliencies, that is, gradients of outputs with regards to inputs, we identify a candidate set of networks that more reliably affect performance in the prediction of general intelligence than similarly sized RSNs. Our approach allows us to further test the effect of local alterations on data and the expected changes in derived metrics such as functional connectivity and instantaneous innovations. Abstract : We predict general intelligence from RS‐fMRI timeseries using a recurrent neural network ensemble in Human Connectome Project data. We propose the selection of networks based on the variance of saliencies per ROI. Resting‐state networks (RSNs) impact on prediction can be explained by their size while with our strategy we find salient networks whose importance exceed that of RSNs.