OBJECT: The goal of this study was to evaluate a novel form of brain surface representation that allows simple, reliable mapping of the surface neuroanatomy for the preoperative evaluation of the spatial relationship between a focal lesion and the precentral gyrus. METHODS: High-resolution three-dimensional (3D) magnetic resonance (MR) imaging data sets were postprocessed using a curved multiplanar reformatting technique to create brain surface reformatted (BSR) images. These BSR images were reconstructed in less than 5 minutes and demonstrated the entire central sulcus with adjacent surface structures in one view. Two experienced neuroradiologists determined the localization of lesions near the central sulcus in 27 patients on standard MR images in three orthogonal planes and on BSR images. In addition, these observers judged whether the lesions were easy or difficult to localize on standard MR and BSR images, and whether diagnoses based on these methods were certain or doubtful. Anatomical localization based on BSR images was compared with that based on functional MR (fMR) images or intraoperative mapping of motor function. The BSR images yielded a perfect concordance with the fMR images and intraoperative mapping (Cohen kappa 1.0) and optimal diagnostic accuracy in localizing perirolandic lesions (both sensitivity and specificity were 100%). Localization was judged to be easy for 48 of 54 diagnoses based on BSR images compared with 26 of 54 based on standard MR images. Diagnoses were assessed as certain for 52 cases based on BSR images and 34 cases based on standard MR images. CONCLUSIONS: Brain surface reformatted imaging improves the diagnostic accuracy of standard anatomical MR imaging for localizing superficial brain lesions in relation to the precentral gyrus. The complementary use of this technique with standard two-dimensional imaging is supported by the fast and simple postprocessing technique and may provide useful information for preoperative surgical planning.
OBJECT: The goal of this study was to evaluate a novel form of brain surface representation that allows simple, reliable mapping of the surface neuroanatomy for the preoperative evaluation of the spatial relationship between a focal lesion and the precentral gyrus. METHODS: High-resolution three-dimensional (3D) magnetic resonance (MR) imaging data sets were postprocessed using a curved multiplanar reformatting technique to create brain surface reformatted (BSR) images. These BSR images were reconstructed in less than 5 minutes and demonstrated the entire central sulcus with adjacent surface structures in one view. Two experienced neuroradiologists determined the localization of lesions near the central sulcus in 27 patients on standard MR images in three orthogonal planes and on BSR images. In addition, these observers judged whether the lesions were easy or difficult to localize on standard MR and BSR images, and whether diagnoses based on these methods were certain or doubtful. Anatomical localization based on BSR images was compared with that based on functional MR (fMR) images or intraoperative mapping of motor function. The BSR images yielded a perfect concordance with the fMR images and intraoperative mapping (Cohen kappa 1.0) and optimal diagnostic accuracy in localizing perirolandic lesions (both sensitivity and specificity were 100%). Localization was judged to be easy for 48 of 54 diagnoses based on BSR images compared with 26 of 54 based on standard MR images. Diagnoses were assessed as certain for 52 cases based on BSR images and 34 cases based on standard MR images. CONCLUSIONS: Brain surface reformatted imaging improves the diagnostic accuracy of standard anatomical MR imaging for localizing superficial brain lesions in relation to the precentral gyrus. The complementary use of this technique with standard two-dimensional imaging is supported by the fast and simple postprocessing technique and may provide useful information for preoperative surgical planning.
Authors: Bruno Fernandes de Oliveira Santos; Daniel de Araujo Paz; Victor Miranda Fernandes; José Calasans Dos Santos; Feres Eduardo Aparecido Chaddad-Neto; Antonio Carlos Sobral Sousa; Joselina Luzia Menezes Oliveira Journal: Sci Rep Date: 2021-03-24 Impact factor: 4.379