Swiss researchers used susceptibility-weighted MRI and an artificial intelligence technique to help identify which individuals with mild cognitive impairment will continue to decline.
Susceptibility-weighted MRI measures magnetic susceptibility phase-effect differences between tissues. Using the technique, researchers were able to view details of the blood vessels in the brain, including microhemorrhages and microbleeds.
The scans revealed microbleeds in 33% subjects with stable mild cognitive impairment. They found microbleeds in 54% of those with progressive cognitive impairment and in 14% of the control group. The study appeared in Radiology.
The researchers also analyzed the data with support vector technology algorithms to identify patterns within a group and create classifications. By adding the support vector technology to the baseline MRI data the researchers were able to distinguish patients with mild cognitive impairment with 85% accuracy.