No abstract

Introduction Advanced machine learning methods might help to identify dementia risk from neuroimaging, but their accuracy to date is unclear. Methods We systematically reviewed the literature, 2006 to late 2016, for machine learning studies differentiating healthy aging from dementia of various types, assessing study quality, and comparing accuracy at different disease boundaries. Results Of 111 relevant studies, most assessed Alzheimer's disease versus healthy controls, using AD Neuroimaging Initiative data, support vector machines, and only T1-weighted sequences. Accuracy was highest for differentiating Alzheimer's disease from healthy controls and poor for differentiating healthy controls versus mild cognitive impairment versus Alzheimer's disease or mild cognitive impairment converters versus nonconverters. Accuracy increased using combined data types, but not by data source, sample size, or machine learning method. Discussion Machine learning does not differentiate clinically relevant disease categories yet. More diverse data sets, combinations of different types of data, and close clinical integration of machine learning would help to advance the field.

Objective:To assess factors associated with white matter hyperintensity (WMH) change in a large cohort after observing obvious WMH shrinkage 1 year after minor stroke in several participants in a longitudinal study.Methods:We recruited participants with minor ischemic stroke and performed clinical assessments and brain MRI. At 1 year, we assessed recurrent cerebrovascular events and dependency and repeated the MRI. We assessed change in WMH volume from baseline to 1 year (normalized to percent intracranial volume [ICV]) and associations with baseline variables, clinical outcomes, and imaging parameters using multivariable analysis of covariance, model of changes, and multinomial logistic regression.Results:Among 190 participants (mean age 65.3 years, range 34.3–96.9 years, 112 [59%] male), WMH decreased in 71 participants by 1 year. At baseline, participants whose WMH decreased had similar WMH volumes but higher blood pressure (p = 0.0064) compared with participants whose WMH increased. At 1 year, participants with WMH decrease (expressed as percent ICV) had larger reductions in blood pressure (β = 0.0053, 95% confidence interval [CI] 0.00099–0.0097 fewer WMH per 1–mm Hg decrease, p = 0.017) and in mean diffusivity in normal-appearing white matter (β = 0.075, 95% CI 0.0025–0.15 fewer WMH per 1-unit mean diffusivity decrease, p = 0.043) than participants with WMH increase; those with WMH increase experienced more recurrent cerebrovascular events (32%, vs 16% with WMH decrease, β = 0.27, 95% CI 0.047–0.50 more WMH per event, p = 0.018).Conclusions:Some WMH may regress after minor stroke, with potentially better clinical and brain tissue outcomes. The role of risk factor control requires verification. Interstitial fluid alterations may account for some WMH reversibility, offering potential intervention targets.