Detecting cerebral microbleeds in 7.0 T MR images using the radial symmetry transform

Cerebral microbleeds have recently received an increased interest, because they appear to be markers of increased risk of vascular events and dementia. Detection and scoring of microbleeds currently requires extensive manual evaluation and hence is very time-consuming. The rating time may be significantly decreased by automated detection of microbleeds using the radial symmetry transform. The goal is to automatically detect cerebral microbleeds in high-resolution MR brain scans, while reducing the number of false positives that need to be removed afterwards by a human rater. A proof of principle experiment was performed and evaluated with two participants of whom cerebral microbleeds were scored by human raters. As an indication of what the proposed method may accomplish, the experiment showed that human rating time reduced from 30 to 1.5 minutes per participant.