gms | German Medical Science

Objective: fMRI is a well-established method to identify functional areas of the brain in awake patients while performing different tasks. Resting state (RS) fMRI enables such localization by detecting synchronous low-frequency blood flow fluctuations in functionally connected brain areas in resting patients performing no tasks. No evidence is available on the RS brain activity in an intraoperative MRI setting with anesthetized patients. Goal of this study was to examine the presence of reproducible resting state (RS) brain activity and functional connectivity patterns of the auditory network during anesthesia in patients with brain tumors in an intraoperative MRI (iMRI).

Method: Twenty subjects with intracranial masses were prospectively studied by means of intraoperative resting-state (RS) fMRI acquisitions prior to craniotomy in an iMRI unit. After performing single-subject independent component analysis (ICA), spatial maps and time courses were assigned to an auditory resting-state network template from the literature and compared via spatial regression coefficients.

Results: All fMRI data were of sufficient quality for further post-processing. In all but one patient the RS functional connectivity of the auditory network could be successfully mapped. In almost all patients with temporal lesions contralateral activation of the auditory network was present. Ipsilaterally to the lesion, activation had a more unstable pattern as it was visualized in 3 out of 5 patients. In patients with tumors in the frontal gyri, the ICA analysis showed contralateral activation of the auditory network in all cases. In all but one patient with tumors in the parietal area, RS-activity of the contralateral auditory network was found.

Conclusions: The auditory cortex was reliably assessable in RS fMRI in an iMRI setting. The combined use of iMRI and RS fMRI in anesthetized patients might be a valuable tool in localizing functional areas of the brain in patients with intracranial lesions, also for intraoperative resection control after brain-shift has occurred. This might influence the surgical strategy to prevent patients from suffering postoperative neurological deficits. Future studies will address the verification of sensorimotor and speech-related RS networks in an intraoperative setting to prepare verification of these analyses even after craniotomy.