208 Connectivity-Based Functional Parcellation and Localization of the Human Supplementary Motor Area Based on Resting-State Functional Magnetic Resting Imaging and Its Utility in Brain Tumor Surgery

• Published in: Neurosurgery Vol. 63; no. CN_suppl_1; pp. 181 - 182
• Main Authors: Zhu, Fengping, Zhuang, Dongxiao, Luo, Qiang, Qiu, Tianming, Wu, Jinsong, Feng, Jianfeng, Mao, Ying
• Format: Journal Article
• Language: English
• Published: United States Oxford University Press 01.08.2016; Copyright by the Congress of Neurological Surgeons; Wolters Kluwer Health, Inc
• Subjects: Brain cancer; Glioma; Localization; Neurosurgery; Surgery
• ISSN: 0148-396X; 1524-4040
• DOI: 10.1227/01.neu.0000489777.45804.bb

Abstract INTRODUCTION: To investigate the utility of resting-state functional MRI (R-fMRI) in the functional parcellation and localization of supplementary motor area (SMA) in healthy subjects and patients with glioma. The utility of this novel technique in brain tumor surgery around SMA was also assessed. METHODS: R-fMRI was acquired from 249 healthy and 8 patients. Defined by AAL template, SMA was parcellated into 3 subregions according to functional connectivity (FC) patterns based on R-fMRI among healthy subjects. The function of each subregion was characterized by assessing whole-brain FC and meta-analysis of existing functional imaging studies. In areas consisting of voxels within and around (2-voxel expansion) SMA, FC was calculated between each voxel and the SMA subregions. Support vector machine was used to discriminate voxels in or outside each subregion. Then, the technique was used to localize the SMA in patients with glioma. The subregions of SMA were precisely stimulated using direct cortical stimulation (DCS) guided by intraoperative MRI when patients do finger opposition motor tasks during surgery. RESULTS: The method has successfully parcellated SMA into 3 subregions, including pre-, mid-, and post-SMA. The function of pre-, mid-, and post-SMA was supposed to be correlated with conflict, eye movement, and hand movement and coordination, respectively. The accuracy of localizing the SMA in the contralateral hemisphere was 87.4% with 1-voxel error by 10-fold cross-validation in the healthy subjects. SMA localization was validated in all patients with DCS. Complete surgical resection was achieved in 6 patients and subtotal in 2 patients. In the short-term period, 3 patients experienced postoperative worsening, however, all recovered to normal at 1-month follow-up. CONCLUSION: Connectivity-based technique based on R-fMRI is effective to parcellate and localize the SMA. Integration of R-fMRI and DCS is useful to identify and protect the SMA in patient-specific surgery for cerebral glioma around SMA.