TY - JOUR T1 - MRI-Based Correction for Partial-Volume Effect Improves Detectability of Intractable Epileptogenic Foci on <sup>123</sup>I-Iomazenil Brain SPECT Images JF - Journal of Nuclear Medicine JO - J Nucl Med SP - 383 LP - 389 DO - 10.2967/jnumed.107.046136 VL - 49 IS - 3 AU - Hiroki Kato AU - Eku Shimosegawa AU - Naohiko Oku AU - Kazuo Kitagawa AU - Haruhiko Kishima AU - Youichi Saitoh AU - Amami Kato AU - Toshiki Yoshimine AU - Jun Hatazawa Y1 - 2008/03/01 UR - http://jnm.snmjournals.org/content/49/3/383.abstract N2 - 123I-Iomazenil brain SPECT has been used for the detection of epileptogenic foci, especially when surgical intervention is considered. Although epileptogenic foci exhibit a decrease in 123I-iomazenil accumulation, normal cerebral cortices often exhibit similar findings because of thin cortical ribbons, gray matter atrophy, or pathologic brain structures. In the present study, we created 123I-iomazenil SPECT images corrected for gray matter volume using MRI and tested whether the detectability of the epileptogenic foci improved. Methods: Seven patients (1 male patient and 6 female patients; mean age ± SD, 34 ± 17 y) with intractable epilepsy were surgically treated by resecting the cerebral cortex after surface electroencephalography. Histopathologic examination of the resected specimens and a good outcome after surgery indicated that the resected lesions were epileptogenic foci. These patients underwent 123I-iomazenil SPECT and 3-dimensional T1-weighted MRI examinations before their operations. Each SPECT image was coregistered to the corresponding MR image, and its partial-volume effect (PVE) was corrected on a voxel-by-voxel basis with a smoothed gray matter distribution image. Four nuclear medicine physicians visually evaluated the 123I-iomazenil SPECT images with and without the PVE correction. The SPECT count ratio of the suspected focus to the contralateral cerebral cortex was evaluated as an asymmetry index (%) based on the volume of interest. Results: The sensitivity, specificity, and accuracy of focus detection by visual assessment were higher after PVE correction (88%, 99%, and 98%, respectively) than before correction (50%, 92%, and 87%, respectively). The mean asymmetry index for the surgically resected lesions was significantly higher on the PVE-corrected SPECT images (22%) than on the PVE-uncorrected ones (16%) (P = 0.006). Conclusion: MRI-based PVE correction for 123I-iomazenil brain SPECT improves the sensitivity and specificity of the detection of cortical epileptogenic foci in patients with intractable epilepsy. ER -