Individual variation underlying brain age estimates in typical development

Abstract

Typical brain development follows a protracted trajectory throughout childhood and adolescence. Deviations from typical growth trajectories have been implicated in neurodevelopmental and psychiatric disorders. Recently, the use of machine learning algorithms to model age as a function of structural or functional brain properties has been used to examine advanced or delayed brain maturation in healthy and clinical populations. Termed 'brain age', this approach often relies on complex, nonlinear models that can be difficult to interpret. In this study, we use model explanation methods to examine the cortical features that contribute to brain age modelling on an individual basis. In a large cohort of n=768 typically-developing children (aged 3-21 years), we build models of brain development using three different machine learning approaches. We employ SHAP, a model-agnostic technique to estimate sample-specific feature importance, to identify regional cortical metrics that explain errors in brain age prediction. We find that, on average, brain age prediction and the cortical features that explain model predictions are consistent across model types and reflect previously reported patterns of regional brain development. However, while several regions are found to contribute to brain age prediction, we find little spatial correspondence between individual estimates of feature importance, even when matched for age, sex and brain age prediction error. We also find no association between brain age error and cognitive performance in this typically-developing sample. Overall, this study shows that, while brain age estimates based on cortical development are relatively robust and consistent across model types and preprocessing strategies, significant between-subject variation exists in the features that explain erroneous brain age predictions on an individual level.

Sex differences in predictors and regional patterns of brain age gap estimates

“…Brain structure shows profound age‐related changes throughout the lifespan (Dima et al, 2022; Frangou et al, 2022; Lebel & Beaulieu, 2011; Storsve et al, 2014; Tamnes et al, 2013; Wierenga et al, 2022), which are also modified by sex. Females show somewhat accelerated brain maturation during adolescence, suggesting a link with pubertal onset (Ball et al, 2021; Brouwer et al, 2021). In middle and late adulthood, sex differences in age‐related brain changes appear less pronounced (Bittner et al, 2021) but female brains may retain more “youthful” transcriptomic and metabolic features than male brains (Beheshti et al, 2021; Berchtold et al, 2008; Goyal et al, 2017, 2019; Skene et al, 2017).…”

Accelerated functional brain aging in major depressive disorder: evidence from a large scale fMRI analysis of Chinese participants

“…Next, we utilize the model to estimate the brain age of 1276 MDD patients on the test set. The brain age prediction is first carried out by three classical supervised learning algorithms: elastic net [38,39], bayesian ridge [40], and ridge regression [30,41,42]. Furthermore, we introduce a stacking model [34] from ensemble learning [43,44] to combine results from the three algorithms, which gives the best estimation results.…”

Investigating Explainable Artificial Intelligence for MRI-based Classification of Dementia: a New Stability Criterion for Explainable Methods

“…It attempts to enhance interpretability by computing the importance values for each feature in the model output [3,4]. Applications of SHAP method are currently present in different domains of research, such as neuroimaging and brain age [5], coronary heart disease [6], chemistry [7] and finance [8]. However, the list of the most informative features might be affected if the predictor variables are highly correlated, possibly resulting into misleading explanations [9,10].…”