USING FUZZY C-MEANS INDEX MATRIX TO VALID THE CORTICAL MICRO-DISORGANIZATION THAT CORRESPONDED TO FOCAL EPILEPTOGENESIS
Abstract number :
1.172
Submission category :
5. Human Imaging
Year :
2009
Submission ID :
9555
Source :
www.aesnet.org
Presentation date :
12/4/2009 12:00:00 AM
Published date :
Aug 26, 2009, 08:12 AM
Authors :
Tsu-Wang Shen and Y. Hsin
Rationale: Segmentation of tissues in medical images is an important image-processing tool for tissue classification, diagnostic purposes, and an initial step for 3D reconstruction from a set of 2D scans. Fuzzy C-means (FCM) is widely used to perform a segmentation process on MR images of the human brain. We incidentally found some unsolved problems in segmentation of brain MR images by using FCM. Blurring of gray-white matter junction was depicted by FCM from patients with focal epilepsy. In this study, we investigated the function of FCM in the identification of cortical micro-disorder and the validation of focal epileptogenesis correspondence. Methods: We retrospectively reviewed the patients who had long-term EEG recording to identify the cortical region with generation of epileptiform discharges at Tzu Chi General Hospital, Hualien, Taiwan. Patients with extra-temporal lobe (neocortical) epilepsy, and with standardized epilepsy-study MR scans were enrolled in the study. Fast SPGR (fast spoiled gradient echo) sequence was used with the following sets of acquisition parameters at GE Signa HD× 1.5-T MR system: TR/TE = 12/5.2 ms, flip angle = 20°, matrix = 512 × 512, and slice thickness = 1.0 mm. For evaluation of neocortical microstructural abnormality, patients with visible MRI abnormality such as mesial temporal lobe sclerosis, hippocampal atrophy or structural abnormality nor spikes arising from sphenoid lead were excluded. For FCM, the power index m is set to 2. Then, the fuzzy indicator matrix I with threshold at 0.6 can show the degree of blurring of the gray-white matter junction. Thereafter we compared the recruited EEGs with the FCM processed images for whether the region of gray-white matter junction blurring and the region of focal spiking were matched. Results: The results showed that micro disorder of cortical organization can be identified successfully according to long-term EEG. The region of cortical micro-disorganization was highly concordant to the location with focal epileptogenic activity. Figure 1 indicates that the region of focal spiking matched the region of the fuzzy-indicator image, but tissue structure changes may not be visible by the regular MRI image. Conclusions: It is not common to find visible MRI abnormality in patients with focal epilepsy, even assisted with the positron emission tomography or the single photon emission computed tomography. Fuzzy C-means is able to valid the cortical micro-disorganization. The technique can help a practitioner of neurology or neurosurgeon to define the focal lesion when an epilepsy surgery is planned to conduct.
Neuroimaging