From a deep learning model back to the brain - inferring morphological markers and their relation to aging

AbstractWe present a Deep Learning framework for the prediction of chronological age from structural MRI scans. Previous findings associate an overestimation of brain age with neurodegenerative diseases and higher mortality rates. However, the importance of brain age prediction goes beyond serving as biomarkers for neurological disorders. Specifically, utilizing convolutional neural network (CNN) analysis to identify brain regions contributing to the prediction can shed light on the complex multivariate process of brain aging. Previous work examined methods to attribute pixel/voxel-wise contributions to the prediction in a single image, resulting in ‘explanation maps’ that were found noisy and unreliable. To address this problem, we developed an inference framework for combining these maps across subjects, thus creating a population-based rather than a subject-specific map. We applied this method to a CNN ensemble trained on predicting subjects’ age from raw T1 brain images of 10,176 subjects. Evaluating the model on an untouched test set resulted in mean absolute error of 3.07 years and a correlation between chronological and predicted age of r=0.98. Using the inference method, we revealed that cavities containing CSF, previously found as general atrophy markers, had the highest contribution for age prediction. Comparing maps derived from different models within the ensemble allowed to assess differences and similarities in brain regions utilized by the model. We showed that this method substantially increased the replicability of explanation maps, converged with results from voxel-based morphometry age studies and highlighted brain regions whose volumetric variability contributed the most to the prediction.HighlightsCNNs ensemble is shown to estimate “brain age” from sMRI with an MAE of ∼3.1 yearsA novel framework enables to highlight brain regions contributing to the predictionThis framework results in explanation maps showing consistency with the literatureAs sample size increases, these maps show higher inter-sample replicabilityCSF cavities reflecting general atrophy were found as a prominent aging biomarker

Download Full-text

Towards the Interpretability of Deep Learning Models for Human Neuroimaging

10.1101/2021.06.25.449906 ◽

2021 ◽

Author(s):

Simon M. Hofmann ◽

Frauke Beyer ◽

Sebastian Lapuschkin ◽

Loeffler Markus ◽

Klaus-Robert Mueller ◽

...

Keyword(s):

Deep Learning ◽

Large Scale ◽

Brain Aging ◽

Accelerated Aging ◽

Population Based ◽

Learning Models ◽

Brain Images ◽

Subarachnoid Spaces ◽

Magnetic Resonance Imaging Mri ◽

Brain Age

Brain-age (BA) estimates based on deep learning are increasingly used as neuroimaging biomarker for brain health; however, the underlying neural features have remained unclear. We combined ensembles of convolutional neural networks with Layer-wise Relevance Propagation (LRP) to detect which brain features contribute to BA. Trained on magnetic resonance imaging (MRI) data of a population-based study (n=2637, 18-82 years), our models estimated age accurately based on single and multiple modalities, regionally restricted and whole-brain images (mean absolute errors 3.38-5.07 years). We find that BA estimates capture aging at both small and large-scale changes, revealing gross enlargements of ventricles and subarachnoid spaces, as well as lesions, iron accumulations and atrophies that appear throughout the brain. Divergence from expected aging reflected cardiovascular risk factors and accelerated aging was more pronounced in the frontal lobe. Applying LRP, our study demonstrates how superior deep learning models detect brain-aging in healthy and at-risk individuals throughout adulthood.

Download Full-text

Contributing Factors to Advanced Brain Aging in Depression and Anxiety Disorders

10.1101/2020.06.16.20132613 ◽

2020 ◽

Author(s):

Laura K.M. Han ◽

Hugo G. Schnack ◽

Rachel M. Brouwer ◽

Dick J. Veltman ◽

Nic J.A. van der Wee ◽

...

Keyword(s):

Brain Aging ◽

Antidepressant Medication ◽

Structural Mri ◽

Depression Symptoms ◽

Depression And Anxiety ◽

Contributing Factors ◽

Major Depressive ◽

Antidepressant Use ◽

Brain Age ◽

Age Prediction

ABSTRACTBrain aging has shown to be more advanced in patients with Major Depressive Disorder (MDD). This study examines which factors underlie this older brain age. Adults aged 18-57 years from the Netherlands Study of Depression and Anxiety underwent structural MRI. A pre-trained brain age prediction model based on >2,000 samples from the ENIGMA consortium was applied to predict age in 65 controls and 220 patients with current MDD and/or anxiety disorder. Brain-predicted age differences (brain-PAD) were calculated (predicted brain age minus chronological age) and associated with clinical, psychological, and biological factors. After correcting for antidepressant use, brain-PAD was significantly higher in MDD (+2.78 years) and anxiety patients (+2.91 years) compared to controls. Findings further indicate unique contributions of higher severity of somatic depression symptoms to advanced brain aging and a potential protective effect of antidepressant medication (-2.53 years).

Download Full-text

Deep learning-based brain age prediction in normal aging and dementia

10.21203/rs.3.rs-804454/v1 ◽

2021 ◽

Author(s):

Jeyeon Lee ◽

Brian Burkett ◽

Hoon-Ki Min ◽

Matthew Senjem ◽

Emily Lundt ◽

...

Keyword(s):

Deep Learning ◽

Cognitive Impairment ◽

Brain Aging ◽

Lewy Body Dementia ◽

Normal Aging ◽

Specific Pattern ◽

Normal Brain ◽

Age Gap ◽

Brain Age ◽

Age Prediction

Abstract Normal brain aging is accompanied by patterns of functional and structural change. Alzheimer's disease (AD), a representative neurodegenerative disease, has been linked to accelerated brain aging at respective age ranges. Here, we developed a deep learning-based brain age prediction model using fluorodeoxyglucose (FDG) PET and structural MRI and tested how the brain age gap relates to degenerative cognitive syndromes including mild cognitive impairment, AD, frontotemporal dementia, and Lewy body dementia. Occlusion analysis, performed to facilitate interpretation of the model, revealed that the model learns an age- and modality-specific pattern of brain aging. The elevated brain age gap in dementia cohorts was highly correlated with the cognitive impairment and AD biomarker. However, regions generating brain age gaps were different for each diagnosis group of which the AD continuum showed similar patterns to normal aging in the CU.

Download Full-text

A reusable benchmark of brain-age prediction from M/EEG resting-state signals

10.1101/2021.12.14.472691 ◽

2021 ◽

Author(s):

Denis A Engemann ◽

Apolline Mellot ◽

Richard Hoechenberger ◽

Hubert Banville ◽

David Sabbagh ◽

...

Keyword(s):

Machine Learning ◽

Deep Learning ◽

Brain Aging ◽

Population Level ◽

Learning Approaches ◽

Eeg Signals ◽

Brain Images ◽

Clinical Endpoints ◽

Starting Point ◽

Brain Age

Population-level modeling can define quantitative measures of individual aging by applying machine learning to large volumes of brain images. These measures of brain age, obtained from the general population, helped characterize disease severity in neurological populations, improving estimates of diagnosis or prognosis. Magnetoencephalography (MEG) and Electroencephalography (EEG) have the potential to further generalize this approach towards prevention and public health by enabling assessments of brain health at large scales in socioeconomically diverse environments. However, more research is needed to define methods that can handle the complexity and diversity of M/EEG signals across diverse real-world contexts. To catalyse this effort, here we propose reusable benchmarks of competing machine learning approaches for brain age modeling. We benchmarked popular classical machine learning pipelines and deep learning architectures previously used for pathology decoding or brain age estimation in 4 international M/EEG cohorts from diverse countries and cultural contexts, including recordings from more than 2500 participants. Our benchmarks were built on top of the M/EEG adaptations of the BIDS standard, providing tools that can be applied with minimal modification on any M/EEG dataset provided in the BIDS format. Our results suggest that, regardless of whether classical machine learning or deep learning was used, the highest performance was reached by pipelines and architectures involving spatially aware representations of the M/EEG signals, leading to R^2 scores between 0.60-0.71. Hand-crafted features paired with random forest regression provided robust benchmarks even in situations in which other approaches failed. Taken together, this set of benchmarks, accompanied by open-source software and high-level Python scripts, can serve as a starting point and quantitative reference for future efforts at developing M/EEG-based measures of brain aging. The generality of the approach renders this benchmark reusable for other related objectives such as modeling specific cognitive variables or clinical endpoints.

Download Full-text

Estimating brain age from structural MRI and MEG data: Insights from dimensionality reduction techniques

10.1101/859660 ◽

2019 ◽

Author(s):

Alba Xifra-Porxas ◽

Arna Ghosh ◽

Georgios D. Mitsis ◽

Marie-Hélène Boudrias

Keyword(s):

Dimensionality Reduction ◽

Brain Aging ◽

Principal Component ◽

Structural Mri ◽

Prediction Performance ◽

Subcortical Structures ◽

Mri Features ◽

Neuroimaging Data ◽

Brain Age ◽

Age Prediction

AbstractBrain age prediction studies aim at reliably estimating the difference between the chronological age of an individual and their predicted age based on neuroimaging data, which has been proposed as an informative measure of disease and cognitive decline. As most previous studies relied exclusively on magnetic resonance imaging (MRI) data, we hereby investigate whether combining structural MRI with functional magnetoencephalography (MEG) information improves age prediction using a large cohort of healthy subjects (N=613, age 18-88 yrs) from the Cam-CAN repository. To this end, we examined the performance of dimensionality reduction and multivariate associative techniques, namely Principal Component Analysis (PCA) and Canonical Correlation Analysis (CCA), to tackle the high dimensionality of neuroimaging data. Using MEG features (mean absolute error (MAE) of 9.60 yrs) yielded worse performance when compared to using MRI features (MAE of 5.33 yrs), but a stacking model combining both feature sets improved age prediction performance (MAE of 4.88 yrs). Furthermore, we found that PCA resulted in inferior performance, whereas CCA in conjunction with Gaussian process regression models yielded the best prediction performance. Notably, CCA allowed us to visualize the features that significantly contributed to brain age prediction. We found that MRI features from subcortical structures were more reliable age predictors than cortical features, and that spectral MEG measures were more reliable than connectivity metrics. Our results provide an insight into the underlying processes that are reflective of brain aging, yielding promise for the identification of reliable biomarkers of neurodegenerative diseases that emerge later during the lifespan.

Download Full-text

Multidimensional brain-age prediction reveals altered brain developmental trajectory in psychiatric disorders

10.1101/2020.12.24.424350 ◽

2020 ◽

Author(s):

Xin Niu ◽

Alexei Taylor ◽

Russell T. Shinohara ◽

John Kounios ◽

Fengqing Zhang

Keyword(s):

Psychiatric Disorders ◽

Robust Regression ◽

Developmental Trajectories ◽

Brain Structures ◽

Brain Regions ◽

Developmental Trajectory ◽

Imaging Features ◽

Neuroimaging Data ◽

Brain Age ◽

Age Prediction

AbstractBrain regions change in different ways and at different rates. This staggered developmental unfolding is determined by genetics and postnatal experience and is implicated in the progression of psychiatric and neurological disorders. Neuroimaging-based brain-age prediction has emerged as an important new approach for studying brain development. However, the unidimensional brain-age estimates provided by previous methods do not capture the divergent developmental trajectories of various brain structures. Here we propose and illustrate an analytic pipeline to compute an index of multidimensional brain-age that provides regional age predictions. First, using a database of 556 subjects that includes psychiatric and neurological patients as well as healthy controls we conducted robust regression to characterize the developmental trajectory of each MRI-based brain-imaging feature. We then utilized cluster analysis to identify subgroups of imaging features with a similar developmental trajectory. For each identified cluster, we obtained a brain-age prediction by applying machine-learning models with imaging features belonging to each cluster. Brain-age predictions from multiple clusters form a multidimensional brain-age index (MBAI). The MBAI is more sensitive to alterations in brain structures and captured distinct regional change patterns. In particular, the MBAI provided a more flexible analysis of brain age across brain regions that revealed changes in specific structures in psychiatric disorders that would otherwise have been combined in a unidimensional brain age prediction. More generally, brain-age prediction using a subset of homogeneous features circumvents the curse of dimensionality in neuroimaging data.

Download Full-text

Increased Brain Age Gap Estimate (BrainAGE) in Young Adults After Premature Birth

Frontiers in Aging Neuroscience ◽

10.3389/fnagi.2021.653365 ◽

2021 ◽

Vol 13 ◽

Author(s):

Dennis M. Hedderich ◽

Aurore Menegaux ◽

Benita Schmitz-Koep ◽

Rachel Nuttall ◽

Juliana Zimmermann ◽

...

Keyword(s):

Premature Birth ◽

Brain Aging ◽

Structural Mri ◽

Full Term ◽

Intracranial Volume ◽

Age Gap ◽

Increased Risk ◽

Aging Rates ◽

Brain Age ◽

Premature Born

Recent evidence suggests increased metabolic and physiologic aging rates in premature-born adults. While the lasting consequences of premature birth on human brain development are known, its impact on brain aging remains unclear. We addressed the question of whether premature birth impacts brain age gap estimates (BrainAGE) using an accurate and robust machine-learning framework based on structural MRI in a large cohort of young premature-born adults (n = 101) and full-term (FT) controls (n = 111). Study participants are part of a geographically defined population study of premature-born individuals, which have been followed longitudinally from birth until young adulthood. We investigated the association between BrainAGE scores and perinatal variables as well as with outcomes of physical (total intracranial volume, TIV) and cognitive development (full-scale IQ, FS-IQ). We found increased BrainAGE in premature-born adults [median (interquartile range) = 1.4 (−1.3–4.7 years)] compared to full-term controls (p = 0.002, Cohen’s d = 0.443), which was associated with low Gestational age (GA), low birth weight (BW), and increased neonatal treatment intensity but not with TIV or FS-IQ. In conclusion, results demonstrate elevated BrainAGE in premature-born adults, suggesting an increased risk for accelerated brain aging in human prematurity.

Download Full-text

Bayesian Optimisation for Neuroimaging Pre-processing in Brain Age Prediction

10.1101/201061 ◽

2017 ◽

Author(s):

Jenessa Lancaster ◽

Romy Lorenz ◽

Rob Leech ◽

James H Cole

Keyword(s):

Absolute Error ◽

Processing Parameters ◽

Training Dataset ◽

Chronological Age ◽

Support Vector ◽

Voxel Size ◽

Smoothing Kernel ◽

And Performance ◽

Brain Age ◽

Age Prediction

AbstractNeuroimaging-based age predictions using machine learning have been shown to relate to cognitive performance, health outcomes and progression of neurodegenerative disease. However, even leading age-prediction algorithms contain measurement error, motivating efforts to improve experimental pipelines. T1-weighted MRI is commonly used for age prediction, and the pre-processing of these scans involves normalisation to a common template and resampling to a common voxel size, followed by spatial smoothing. Resampling parameters are often selected arbitrarily. Here, we sought to improve brain-age prediction accuracy by optimising resampling parameters using Bayesian optimisation.Using data on N=2001 healthy individuals (aged 16-90 years) we trained support vector machines to i) distinguish between young (<50 years) and old (>50 years) brains and ii) predict chronological age, with accuracy assessed using cross-validation. We also evaluated model generalisability to the Cam-CAN dataset (N=648, aged 18-88 years). Bayesian optimisation was used to identify optimal voxel size and smoothing kernel size for each task. This procedure adaptively samples the parameter space to evaluate accuracy across a range of possible parameters, using independent sub-samples to iteratively assess different parameter combinations to arrive at optimal values.When distinguishing between young and old brains a classification accuracy of 96.25% was achieved, with voxel size = 11.5mm3 and smoothing kernel = 2.3mm. For predicting chronological age, a mean absolute error (MAE) of 5.08 years was achieved, with voxel size = 3.73mm3 and smoothing kernel = 3.68mm. This was compared to performance using default values of 1.5mm3 and 4mm respectively, which gave a MAE = 5.48 years, a 7.3% improvement. When assessing generalisability, best performance was achieved when applying the entire Bayesian optimisation framework to the new dataset, out-performing the parameters optimised for the initial training dataset.Our study demonstrates the proof-of-principle that neuroimaging models for brain age prediction can be improved by using Bayesian optimisation to select more appropriate pre-processing parameters. Our results suggest that different parameters are selected and performance improves when optimisation is conducted in specific contexts. This motivates use of optimisation techniques at many different points during the experimental process, which may result in improved statistical sensitivity and reduce opportunities for experimenter-led bias.

Download Full-text

Brain age prediction of healthy subjects on anatomic MRI with deep learning: going beyond with an "explainable AI" mindset

10.1101/413302 ◽

2018 ◽

Cited By ~ 2

Author(s):

Paul Herent ◽

Simon Jegou ◽

Gilles Wainrib ◽

Thomas Clozel

Keyword(s):

Machine Learning ◽

Deep Learning ◽

Grey Matter ◽

Gray Matter Volume ◽

Brain Regions ◽

Gradient Boosting ◽

Tissue Segmentation ◽

Caudate Nuclei ◽

Skull Stripping ◽

Brain Age

Objectives: Define a clinically usable preprocessing pipeline for MRI data. Predict brain age using various machine learning and deep learning algorithms. Define Caveat against common machine learning traps. Data and Methods: We used 1597 open-access T1 weighted MRI from 24 hospitals. Preprocessing consisted in applying: N4 bias field correction, registration to MNI152 space, white and grey stripe intensity normalization, skull stripping and brain tissue segmentation. Prediction of brain age was done with growing complexity of data input (histograms, grey matter from segmented MRI, raw data) and models for training (linear models, non linear model such as gradient boosting over decision trees, and 2D and 3D convolutional neural networks). Work on interpretability consisted in (i) proceeding on basic data visualization like correlations maps between age and voxels value, and generating (ii) weights maps of simpler models, (iii) heatmap from CNNs model with occlusion method. Results: Processing time seemed feasible in a radiological workflow: 5 min for one 3D T1 MRI. We found a significant correlation between age and gray matter volume with a correlation r = -0.74. Our best model obtained a mean absolute error of 3.60 years, with fine tuned convolution neural network (CNN) pretrained on ImageNet. We carefully analyzed and interpreted the center effect. Our work on interpretability on simpler models permitted to observe heterogeneity of prediction depending on brain regions known for being involved in ageing (grey matter, ventricles). Occlusion method of CNN showed the importance of Insula and deep grey matter (thalami, caudate nuclei) in predictions. Conclusions: Predicting the brain age using deep learning could be a standardized metric usable in daily neuroradiological reports. An explainable algorithm gives more confidence and acceptability for its use in practice. More clinical studies using this new quantitative biomarker in neurological diseases will show how to use it at its best.

Download Full-text

History of childbirths relates to region-specific brain aging patterns in middle and older-aged women

10.1101/2020.05.08.084616 ◽

2020 ◽

Cited By ~ 1

Author(s):

Ann-Marie G. de Lange ◽

Claudia Barth ◽

Tobias Kaufmann ◽

Melis Anatürk ◽

Sana Suri ◽

...

Keyword(s):

Brain Aging ◽

Reward System ◽

Regional Brain ◽

Pregnancy And Postpartum ◽

Maternal Brain ◽

History Of ◽

Novel Applications ◽

Prospective Longitudinal ◽

Brain Age ◽

Age Prediction

AbstractPregnancy involves maternal brain adaptations, but little is known about how parity influences women’s brain aging trajectories later in life. In this study, we replicated previous findings showing less apparent brain aging in women with a history of childbirths, and identified regional brain aging patterns linked to parity in 19,787 middle and older-aged women. Using novel applications of brain-age prediction methods, we found that a higher number of previous childbirths was linked to less apparent brain aging in striatal and limbic regions. The strongest effect was found in the accumbens – a key region in the mesolimbic reward system, which plays an important role in maternal behavior. While only prospective longitudinal studies would be conclusive, our findings indicate that subcortical brain modulations during pregnancy and postpartum may be traceable decades after childbirth.

Download Full-text