deficits . Interestingly, patients with severe peri-lesional edema consistently showed a higher prediction error. This may be related to impairment of vascular auto regulation with decreased activation-dependent changes in cerebral blood flow in regions of high edema. It was also noted that distortion of the fMRI images could be a problem when the region of interest is closely related to bone, as may occur with the inferior temporal lobe.
Identification and mapping of white matter tracts is now also possible . Image-fusion software then allows the accurate fusion of these maps with the anatomical image data set. The presence of the axonal membrane and the neu-rofilamentary cytoskeleton restricts the diffusion of water to the long axis of the fiber tracts - anisotropy. Diffusion-anisotropy MRI identifies the restricted diffusion as a hyperintense area, delineating in three dimensions the position and direction of the tract. In four patients presenting with tumors displacing the pyramidal tract, this technique was helpful in the perioperative identification of the tract. As in fMRI, peri-lesional edema poses a problem, as there is no restriction of diffusion direction in edematous foci; tract identification is then difficult. It has been suggested that this method can also be used to map the optic radiation and the com-missural fibers if these structures are relevant to the procedure.
A critical limitation of image-guided surgery, particularly relevant to tumor resection, is the reliance on a pre-operatively acquired image data set. Progressive movement or distortion of the brain during the procedure means that the pre-operative image becomes progressively outdated as the surgery proceeds . Reliance on the image-guided system would lead to errors in the intraoperative delineation of tumor location and borders, as well as in the relationship of the tumor to adjacent eloquent cortex. Brain distortion, relative to the pre-operative image, is due to the release of cerebrospinal fluid (CSF), pressure changes on skull opening, unopposed gravity, ventricular compression, brain retraction and tumor resection. Patient positioning, the physiological effects of diuretics and mechanical ventilation are also relevant.
Dorward et al.  studied the magnitude and direction of post-imaging brain distortion in 48 cases; maximal brain shifts were found to be greater than 1 cm in magnitude. The degree of shift depended on the size of the tumor, the degree of midline shift, and the presence of peri-tumoral edema; significant differences were identified between meningiomas, gliomas, non-glial intra-axial lesions and skull base lesions.
Nabavi et al.  studied post-imaging brain deformation in 25 patients using a 0.5T vertically open bore MR imager. Baseline imaging was performed after positioning; further images were taken after dural opening and initial CSF drainage, after tumor resection, and after dural closure. Software allowed image overlapping on the same coordinates to facilitate comparison between the images. In this study, brain shift was shown to have a high degree of interindividual variability and was a continuous and dynamic process, evolving separately and differently in distinct brain regions. It was noted that the direction of deformation might even reverse within relatively short time-frames. The occipital and parietal lobes were seen to be less mobile than the frontal and temporal lobes. It is clear that brain biomechanics are still far from understood, and although software may be designed to predict unidirectional surface shifts secondary to gravity and CSF loss associated with small surface lesions, predictions of multidirectional subsurface movements associated with larger lesions are unlikely to be accurate.
The effects of brain shift can be minimized by paying attention to some operative details, such as allowing as little CSF as possible to escape, by delaying aspiration of cystic tumor components, and by removing tumor adjacent to eloquent cortex first. The use of dehydrating agents such as mannitol should be avoided and the positioning of the patient should be such that the craniotomy site lies at the highest point.
One solution to the problem of intraoperative brain movement is to combine guidance from registered pre-operative images with intraoperative MRI. This allows the acquisition of image updates intraoperatively; these updates are then fused with the pre-operatively acquired image data set. Several centers have reported their experience with such systems [17,18,19]. The
Was this article helpful?
The term vaginitis is one that is applied to any inflammation or infection of the vagina, and there are many different conditions that are categorized together under this ‘broad’ heading, including bacterial vaginosis, trichomoniasis and non-infectious vaginitis.