Deep neural networks learn general and clinically relevant representations of the ageing brain

Leonardsen, Esten; Han, Peng; Kaufmann, Tobias; Agartz, Ingrid; Andreassen, Ole A.; Celius, Elisabeth Gulowsen; Espeseth, Thomas; Harbo, Hanne F.; Høgestøl, Einar August; Lange, Ann-Marie G. de; Marquand, André F.; Vidal-Piñeiro, Dídac; Roe, James M; Selbæk, Geir; Sørensen, Øystein; Smith, Stephen M.; Westlye, Lars T.; Wolfers, Thomas; Wang, Yunpeng

doi:10.1016/j.neuroimage.2022.119210

Cited by 74 publications

(141 citation statements)

References 68 publications

Supporting

Mentioning

107

Contrasting

Order By: Relevance

“…While in our study we achieved prediction accuracy higher than in the rest of the published EEG literature, MRI-based brain age prediction of MAE is significantly higher. In a recent study, Leonardsen and colleagues (Leonardsen et al, 2022) achieved MAE = 2 47 . years.…”

Section: Accurate Brain Age Prediction From Eeg Is Feasiblementioning

confidence: 90%

Predicting age from resting-state scalp EEG signals with deep convolutional neural networks on TD-brain dataset

Khayretdinova¹,

Shovkun²,

Degtyarev³

et al. 2022

Front. Aging Neurosci.

View full text Add to dashboard Cite

IntroductionBrain age prediction has been shown to be clinically relevant, with errors in its prediction associated with various psychiatric and neurological conditions. While the prediction from structural and functional magnetic resonance imaging data has been feasible with high accuracy, whether the same results can be achieved with electroencephalography is unclear.MethodsThe current study aimed to create a new deep learning solution for brain age prediction using raw resting-state scalp EEG. To this end, we utilized the TD-BRAIN dataset, including 1,274 subjects (both healthy controls and individuals with various psychiatric disorders, with a total of 1,335 recording sessions). To achieve the best age prediction, we used data augmentation techniques to increase the diversity of the training set and developed a deep convolutional neural network model.ResultsThe model’s training was done with 10-fold cross-subject cross-validation, with the EEG recordings of the subjects used for training not considered to test the model. In training, using the relative rather than the absolute loss function led to a better mean absolute error of 5.96 years in cross-validation. We found that the best performance could be achieved when both eyes-open and eyes-closed states are used simultaneously. The frontocentral electrodes played the most important role in age prediction.DiscussionThe architecture and training method of the proposed deep convolutional neural networks (DCNN) improve state-of-the-art metrics in the age prediction task using raw resting-state EEG data by 13%. Given that brain age prediction might be a potential biomarker of numerous brain diseases, inexpensive and precise EEG-based estimation of brain age will be in demand for clinical practice.

show abstract

Section: Accurate Brain Age Prediction From Eeg Is Feasiblementioning

confidence: 90%

Predicting age from resting-state scalp EEG signals with deep convolutional neural networks on TD-brain dataset

Khayretdinova¹,

Shovkun²,

Degtyarev³

et al. 2022

Front. Aging Neurosci.

View full text Add to dashboard Cite

show abstract

“…We used age as a continuous variable rather than a one-hot encoded matrix (i.e. which would effectively treat the regression as a classification problem 60 ). This enables us to generalize beyond the age range used in the training dataset, which is important for transfer learning because of potential differences between cohorts.…”

Section: Methodsmentioning

confidence: 99%

Explanatory latent representation of heterogeneous spatial maps of task-fMRI in large-scale datasets

Zabihi

Wolfers

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

The increasing number of neuroimaging scans in recent years has facilitated the use of complex nonlinear approaches to analyzing such data. More specifically, deep learning, which has been previously hindered by the curse of dimensionality is now feasible. However, it remains challenging to use these techniques develop reliable biomarkers and find an optimal representation of data that explains the biological underpinnings of the mental disorders Here, we employed a 3-dimensional autoencoder with an architecture designed from the ground up for task-fMRI data. Our study presented a coherent strategy for optimizing model parameters and architecture and a method for visualizing and interpreting the latent space representation. We trained our model with multi-task fMRI data derived from the Human Connectome Project (HCP) that provides whole-brain coverage across a range of cognitive tasks. Next, in a transfer learning setting, we tested the generalization of our latent space on UK Biobank data as an independent dataset. We showed that the model did not only learn salient features such as age but also high-level behavioral characteristics and that this representation was highly generic and generalizable to an independent dataset. Furthermore, we demonstrated that the projection of latent space back into the original space is meaningful and interpretable. Finally, our results show that with careful implementation, nonlinear features can provide complementary information that accessible to purely linear methods. Our results provide an important step toward learning interpretable and generalizable latent representations that link cognition with underlying brain systems

show abstract

“…State-of-the-art conceptualizations of health, such as the bio-psycho-social model (Engel, 1977), recommend considering various domains or levels of explanation when assessing health outcomes, such as brain age. To date, brain age has usually been described from single variables while controlling for sex and age (e.g., Cole, 2020;Leonardsen et al, 2022). However, cumulative and synergy effects can be expected to partly explain health, which has, for example, been shown for cardiometabolic risk factors explaining brain age (see Beck et al, 2022aBeck et al, , 2022b.…”

Section: (Which Was Not Certified By Peer Review) Preprintmentioning

confidence: 99%

Bio-psycho-social factors’ associations with brain age: a large-scale UK Biobank diffusion study of 35,749 participants

Korbmacher

Gurholt

Lange

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Brain age refers to age predicted by brain features. Brain age has previously been associated with various health and disease outcomes and suggested as a potential biomarker of general health. Few previous studies have systematically assessed brain age variability derived from single and multi-shell diffusion magnetic resonance imaging data. Here, we present multivariate models of brain age derived from various diffusion approaches and how they relate to bio-psycho-social variables within the domains of sociodemographic, cognitive, life-satisfaction, as well as health and lifestyle factors in midlife to old age (N = 35,749, 44.6 to 82.8 years of age). Bio-psycho-social factors could uniquely explain only a small proportion of the brain age variance. We observed a similar pattern across diffusion approaches. We consistently identified bio-psycho-social associations of interest across models. Waist-to-hip ratio, diabetes, hypertension, smoking, matrix puzzles solving, and job and health satisfaction and perception were associated with brain age. At the same time, sex and ethnicity showed large variability in their predictor strengths. As expected, chronological age contributed the strongest to explaining variability in brain age. Our results show that brain age cannot be sufficiently explained by bio-psycho-social variables alone. However, the observed associations suggest to adjust for sex, ethnicity, cognitive factors, as well as health and lifestyle factors.

show abstract

Deep neural networks learn general and clinically relevant representations of the ageing brain

Cited by 74 publications

References 68 publications

Predicting age from resting-state scalp EEG signals with deep convolutional neural networks on TD-brain dataset

Predicting age from resting-state scalp EEG signals with deep convolutional neural networks on TD-brain dataset

Explanatory latent representation of heterogeneous spatial maps of task-fMRI in large-scale datasets

Bio-psycho-social factors’ associations with brain age: a large-scale UK Biobank diffusion study of 35,749 participants

Contact Info

Product

Resources

About