MRI at 3 T of brain functions and fibre tracts adjacent to intracranial tumors

University dissertation from Diagnostisk radiologi, Lunds Universitet

Abstract: The aims of the studies were: to evaluate the spatial resolution of fMRI at 3 T by studying fingersomatotopy in area 3 b of the primary sensory area (Paper 1); to investigate the potential of fMRI at 3T as a clinical tool for preoperative evaluation of patients with intracranial tumors (Paper 2); to investigate the effect of glioma resection on the spatial extent of fMRI activation as compared to the normal within-subject variability (Paper 3); to determine whether the apparent diffusion coefficient (ADC) and fractional anisotropy (FA) can distinguish tumor-infiltrated edema of gliomas from pure edema of meningiomas and metastases (Paper 4). Results: Strict somatotopic organisation in the primary sensory cortex was shown in the group average; at the subject level the thumb was located laterally, anteriorly and inferiorly to the little finger in 94 % of subjects (Paper 1). The sensorimotor and language areas close to intracranial tumors were identified in 95 % of patients. Paradigm effectiveness ranged from 79 to 95 %. The median quality of the activation maps was high. fMRI contributed to the decision to operate, the surgical approach and the extent of the resection in 9, 13 and 12 patients, respectively (Paper 2). BOLD activation obtained from fMRI with motor, sensory and language stimulation pre- and postoperatively in glioma patients yielded differences in the spatial extent similar to those obtained from repeated examinations in healthy controls (Paper 3). Values and lesion-to-brain ratios of ADC and FA in peritumoral edema did not differ between high grade gliomas, meningiomas and metastases (Paper 4). Conclusions: Our results suggest that the spatial resolution of fMRI using a clinical protocol is adequate for localisation of the sensory representation of a finger (Paper 1). Preoperative fMRI for mapping of motor, sensory and language functions at 3 T is feasible and contributes to neurosurgical decision making (Paper 2). Longitudinal pre- and postoperative fMRI studies may be performed in patients with gliomas without concerns for potential effects of the decreased tumor volume on the BOLD effect as a possible source of error (Paper 3). Values and lesion-to-brain ratios of ADC and FA may not be useful to distinguish pure edema from tumor-infiltrated edema in peritumoral areas with T2-signal changes (Paper 4).