MULTI-MODALITY IMAGE REGISTRATION FOR SUBDURAL ELECTRODE LOCALIZATION

摘要: Surgical treatment has been proved to be an effective way control seizures for some kinds of intractable epilepsy. The electrocorticogram (ECoG) recorded from subdural electrodes become important technique defining epileptogenic zones before surgery in clinical practice. exact location determined establish the connection between and zones. Artifacts caused by can severely affect quality CT imaging sequentially image registration. In this paper, we discussed performance mean squares Mattes mutual information metric methods multimodal registration electrode localization. Since skull regarded as a rigid body, is sufficient purpose vital parameter rotation. translation result depends on Both performed well determination rotation center. Rotation angles different pairs same volume pair fluctuated lot. Based acquisition process, assume that images within should have transformation parameters Results show one dataset are approximate manual results considered actual despite their fluctuation range.