Characteristics of perivascular space dilatation in normal aging

Abstract

The increased incidence of dilated perivascular spaces (dPVSs) visible on MRI has been observed with advancing age, but the relevance of PVS dilatation to normal aging across the lifespan has yet to be fully clarified. In the current study, we sought to find out the age dependence of dPVSs by exploring changes in different characteristics of PVS dilatation across a wide range of age. For 1220 healthy subjects aged between 18 and 100 years, PVSs were automatically segmented and characteristics of PVS dilatation were assessed in terms of the burden, location, and morphology of PVSs in the white matter (WM) and basal ganglia (BG). A machine learning model using the random forests method was constructed to estimate the subjects' age by employing the PVS features. The constructed machine learning model was able to estimate the age of the subjects with an error of 9.53 years on average (correlation = 0.875). The importance of the PVS features indicated the primary contribution of the burden of PVSs in the BG and the additional contribution of locational and morphological changes of PVSs, specifically peripheral extension and reduced linearity, in the WM to age estimation. Indeed, adding the PVS location or morphology features to the PVS burden features provided an improvement to the performance of age estimation. The age dependence of dPVSs in terms of such various characteristics of PVS dilatation in healthy subjects could provide a more comprehensive reference for detecting brain disease-related PVS dilatation.

Keywords: age estimation; machine learning; normal aging; perivascular space.

FIGURE 1

Maps of voxel‐wise frequency of detecting perivascular spaces across subjects in each of three age groups from young to middle‐aged to elderly subjects.

FIGURE 2

Relationship between actual age and estimated age for 1220 subjects. While the ideal fit corresponds to perfect estimations of age for all subjects, the empirical fit indicates a linear fit to estimated age from a machine learning model based on perivascular space features. A dot and an errorbar at each subject's data point represent the mean and standard deviation of age estimates over 100 repetitions of five‐fold cross‐validation.

FIGURE 3

Univariate importance of perivascular space (PVS) features. It represents a t value for the corresponding coefficient in a linear regression model of subjects’ age fit to each PVS feature in (a), and that of statistical significance is exhibited in (b). BG, basal ganglia.

FIGURE 4

Multivariate importance of perivascular space (PVS) features. It indicates the relative influence of each PVS feature on the performance of a machine learning model that considered all PVS features simultaneously for the estimation of subjects’ age in (a), and its average over three PVS feature types and five regions is shown in (b). BG, basal ganglia.

FIGURE 5

Comparison of the estimation performance of machine learning models based on different combinations of perivascular space features. The performance of estimations was compared in terms of the root mean squared error (RMSE) of estimated age relative to actual age in (a), and in terms of the correlation between estimated age and actual age in (b). Lower values in the RMSE and higher values in the correlation lead to higher orders in the estimation performance.