A Statistical Approach to the Segmentation ofMR Imagery and Volume Estimation of StrokeLesions

摘要

We propose a three-dimensional method to segment magnetic resonance imagery (MRI)of iscliemic stroke patients into lesion and background and hence to estimate lesion volumes. It is a hierarchical. Regularized method beased on classical statistics that produces confidence regions for the lesion itself and a two-sided confidence interval for lesion volume. Thus approach requires a limited amount of user interaction to initialize. The procedure has been tested on real MR. Data, with volume estimates within 10% of those deived from doctors' hand segmentations. According to the physicians with whom we are working. These results are clincally useful to evaluate stroke therapies.