| Part 1BOLD-fMRI study of the cortical organization of language function in normal subjectsObjective: By studying cortical organization of language function in normal subjects using three different language tasks to find a protocol which can being proper used in presurgical language mapping with BOLD-fMRI. Methods: 10 healthy volunteers with right handiness underwent BOLD-fMRI scan using three different language tasks: picture-naming, word-generation and over learned poem reciting. The functional data were analyzed with AFNI software, which is designed for neuroimaging data processing and Func Tool 1.9 software accompanying with GE station. The activation areas of each task were studied, in order to find which one is more reliable in presurgical use of language mapping. The result analyzed with AFNI software and Func Tool 1.9 software were compared to adjust whether Func Tool 1.9 software can be used as a quick analyzing tool in clinic use. Proper analyzing threshold was concluded.Results: By comparing the activation pattern during three language tasks, we found that the three tasks shared a common brain network, which was dispersed in the frontal, parietal, temporal, and occipital lobe. We also found that the activation during word generation task were more consistent and strong than during other two tasks. Furthermore, we found that the activation patterns for picture naming and over learned poem reciting were bilateral and that for word generation was mainly left lateralized. The activation areas displayed with Func Tool 1.9 software analyzing were just accordant with those with AFNI software analyzing. A threshold of p=0.001, with T value between 0.5 to 0.6 was reliable for individual analyzing of most subjects in different tasks.Conclusion: In addition of good lateralizing value, word generation was also an useful task in localizing language areas in both frontal lobe and tempoparietal lobe. Func Tool 1.9 software can be used as a quick analyzing method in prediction of language areas.Part 2BOLD-fMRI in Preoperative Assessment of Language Areas: Correlation with Direct Cortical Stimulation.Objective: The aim of this study was to analyze the usefulness of preoperative language functional magnetic resonance imaging (fMRI), by correlating fMRI data with intraoperative cortical stimulation results for patients with brain tumors. Methods: Word generation task was used for 12 right-handed patients with lesions in the left hemisphere. fMRI data obtained were analyzed with Func Tool software, with analysis thresholds p=0.001.The fMRI data were then registered in a frameless stereotactic neuronavigational device and correlated with direct brain mapping results. We used a statistical model with the fMRI information as a predictor, spatially correlating each intraoperatively mapped cortical site with fMRI data integrated in the neuronavigational system (site-by-site correlation). Two matching criteria (overlapping and adjacent within lcm) were used to adjust the relationship between area actived by language task in BOLD-fMRI and site inducing language interruption in intraoperative cortical simulation.Results: The activated areas were located mainly in the middle and inferior frontal gyri (F2 and F3), the superior and middle temporal gyri (Tl and T2), and the supramarginal and angular gyri, primary motor area (PMA) supplementary motor area (SMA) , insular , occipital lobe and cerebellum. All patients had right hemisphere activations with a less degree. A total of 145 cortical sites were tested among the 12 patients. Poor sensitivity of the fMRI technique was observed by using strict criteria(overlapping within two methods) in total areas(48.1%).while loosening the criteria to adjacent within lcm,the sensitivity was increased to 88.9%, though the specificity was slightly decreased(83.1% to 75%). Although not perfect, thesensitivity and specificity of the fMRI technique was good in benign lesion and low grade tumor (95.2% and 80.4%, respectively). BOLD-fMRI was less reliable in localizing language area in malignant tumors (with sensitivity of 66.7%and specificity of 68.6% respectively).Conclusion: With the paradigms and analysis thresholds used in this study, language fMRI data obtained by word generation task, though were imperfectly correlated with intraoperative brain mapping results using strict criteria, but a better correlation could be obtained by loosening the criteria to adjacent within Icm. Preoperative language BOLD-fMRI may useful in identification language area in patients underwent neurosurgery, especially in those with benign and low grade tumor.Part 3Integrating BOLD functional MR Imaging in neuronavigation surgery in language cortexObjective: In this study, we applied the blood oxygen level dependent (BOLD) functional magnetic resonance imaging (fMRI) for the functional localization of language cortex. The BOLD imaging data was integrated to a neuronavigation system, and proposed as guides for the resection of brain tumors located in langage cortex. A control study was carried out to evaluate the role of BOLD imaging in neuronavigation surgery of brain tumors located in language cortex. Methods: A total of 52 patients with brain lesion adjacent to the language cortex in left cerebral region were defined into trial group(BOLD navigation with or without electrocortical stimulation mapping) or control group (traditional navigation). Therewere 33 male and 19 female cases. Their mean age was 38.2±16.4. 48 cases were right handed, while 4 cases were left or bilateral handed. There were 27 lesions located in frontal lobe, 17 lesions in temporoparietal lobe and 8 lesions involving both frontal temporal areas. According to presurgical language assessment, 44 cases were of normal language skill, while 8 cases were with slight language deficit. For the patients in trial group, word generation task was used in presurgical BOLD-FMRI examination. The anatomical structure was delineated by the Tl-weighted three-dimensional fast spoiled gradient recalled sequence (3D/FSPGR) immediately afterward. The BOLD images were overlay on the Tl-weighted images, and then coregistered to the neuronavigation system. For the cases in control group underwent Tl weighted 3D navigational MR imaging examinations only. The percentage of totally resection was compared within two groups. The language score were examined before and after operation. The long-tem outcome was followed referring to the Karnofsky scale 3 to 6 months postoperatively.Results: The BOLD imaging was successful achieved in 25 of 30 cases. The activated areas were located mainly in the in the middle and inferior frontal gyri (F2 and F3), the superior and middle temporal gyri (Tl and T2), and the supramarginal and angular gyri, primary motor area (PMA), supplementary motor area (SMA) , insular, occipital lobe and cerebellum. All patients had right hemisphere activations in a less degree. All cases were adjusted as left language domination by BOLD activation irregardless with the side of handiness. There were finally 25 cases in trial group(ESM+BOLD and BOLD), and 27 cases in control group respectively. The statistic analysis confirmed well balance of main variations in two groups. The lesions were completely resected in 18 cases of trial group (72%) and 12 cases of control group (44.4%), (P<0.05). Long term follow up language worsen after 3 months were observed in 8% cases of trial group, while 29.6% cases in control group (P<0.05). Postoperative language score decline was fewer in trial group than in control group,... |
PDF Full Text Request