Rapid and accurate anatomical localization of implanted subdural electrodes in a virtual reality environment.

BACKGROUND: An accurate and rapid anatomical localization of implanted subdural electrodes is essential in the invasive diagnostic process for epilepsy surgery. OBJECTIVE: To demonstrate our experience with a three-dimensional (3D) virtual reality simulation software (Dextroscope®, Bracco Imaging, Milano, Italy) in the postoperative localization of subdural electrodes. METHODS: Postoperative thin-slice computed tomography (CT) scans were coregistered to preoperative 3D magnetic resonance (MR) images in the Dextroscope environment in 10 patients. Single-electrode contacts were segmented and their positions in relation to specific brain anatomic structures were obtained by 3D reconstruction within the Dextroscope environment. The spatial accuracy was tested by comparing the positions of the electrode contacts as visible in the 3D reconstruction with intraoperative photographs. Image processing time was also recorded. RESULTS: The 3D stereoscopic reconstruction provided an accurate representation of the implanted electrodes with highly detailed visualization of the underlying anatomy. The mean absolute difference between 3D reconstruction and intraoperative photographs was 2.4 mm ± 2.2 mm. The processing time to obtain the 3D reconstructions did not exceed 15 minutes. CONCLUSIONS: The results indicate that the 3D virtual reality simulation software used in our series is a useful tool for rapid and precise localization of subdural electrodes implanted for invasive electroencephalography (EEG) recordings.

The minimally invasive supraorbital subfrontal key-hole approach for surgical treatment of temporomesial lesions of the dominant hemisphere.

INTRODUCTION: Surgery in the temporomesial region is generally performed using a subtemporal, transtemporal, or pterional-transsylvian approach. However, these approaches may lead to approach-related trauma of the temporal lobe and frontotemporal operculum with subsequent postoperative neurological deficits. Iatrogenic traumatisation is especially significant if surgery is performed in the dominant hemisphere. METHODS: During a five-year period between January 2003 and December 2007, we have approached the temporomesial region in 21 cases via the supraorbital approach. In 15 cases, the lesion was located within the dominant hemisphere, all lesions had space-occupying effects. In all cases, meticulous approach planning was performed, demonstrating a close proximity of the lesion to the pial surface on the upper anterior mesial aspect of the temporal lobe. An extension within the parahippocampal gyrus or with deep temporobasal tumor growth below the sphenoid wing were considered as exclusion criteria for using the supraorbital approach. RESULTS: In all cases surgery was performed without intraoperative complications. Pathological investigation showed 7 low-grade astrocytomas, 4 high-grade astrocytomas, 2 gangliogliomas and 2 cavernomas. Early postoperative MRI scans confirmed a complete removal of the lesion in 14 cases. In one case of a subtotal resection, the residual tumor was removed through a posterior subtemporal approach. The postoperative neurological examination was unchanged in 14 cases. In one case a transient hemiparesis was observed. In patients with dominant-sided lesions no speech or mental deficits were present. CONCLUSION: In selected cases, the minimally invasive supraorbital craniotomy offers excellent surgical efficiency in the temporomesial region with no approach-related morbidity compared to a standard transtemporal or pterional-transsylvian approach.

Minimally invasive cerebral cavernoma surgery using keyhole approaches - solutions for technique-related limitations.

Cavernomas are often small in size and located in difficultly accessible regions. Preoperative identification of the ideal surgical approach as well as the precise intraoperative implementation of the surgical plan are of critical importance for successful surgery. While aiming for minimally invasive surgical techniques and maximally effective cavernoma resection, we envisaged that employing a combination of precise and technically sophisticated virtual reality surgery planning, modern navigation systems with augmented reality features and endoscope-assisted surgical techniques should contribute to achieve this goal. Between December 2002 and November 2005, 66 patients were operated on for cerebral cavernomas in our department. In 23 cases surgery planning was done by using a virtual reality planning system, neuronavigation was used in 43 cases and the intraoperative augmented reality feature was used in 16 cases. 10 patients were operated by using the endoscopic assisted surgical technique. Complete resection was achieved in all cases. Using all nowadays available surgical tools, cerebral cavernomas can be operated with minimally invasive techniques and with excellent results.

A new model of skull base reconstruction following expanded endonasal or transoral approaches--long-term results in primates.

OBJECTIVE: The direct endonasal or transoral transclival approaches to the skull base permit effective minimally invasive surgery along the clivus region. Developing consistently effective techniques to prevent cerebrospinal fluid (CSF) leaks and their consequences (infections and healing processes with long and complicated recoveries) remains a major challenge. In this study, we tested over a long period a method of bone reconstruction newly developed by us, which makes use of a specially designed elastic silicone plug that can be employed for bone replacement after minimally invasive skull base surgery without risk of postoperative CSF leaks. After acute testing of plug efficiency in a pig model, which showed a 100% closure of the bone defect without CSF leak, we now tested the long-term accuracy of the plugs. METHODS: In 3 primates, we used an endoscope-controlled transoral transclival approach and after opening the dura we simulated a CSF leakage. We inserted the plug into the bone defect and closed the mucosa of the oral cavity with stitches. The follow-up included blood, weight, and wound control 1, 4 and 8 weeks postoperatively. Social behavior, such as reintegration and postoperative eating abnormalities, was also studied. The aims of this study were: (1) testing the biocompatibility of the material; (2) development of infection against the foreign body; (3) effects of the plug on the surrounding bone, and (4) development of CSF leakages during the postoperative phase. RESULTS: Clinically no infection was seen. Wound healing, immediate and long-term postoperative social behavior of the animals, feeding and body weight were normal. No CSF leakages developed. The histological examination of the clivus bone showed no abnormalities. The implant was covered by fibrous layer; there was no bone atrophy but osteoid formation. CONCLUSION: This novel medical device allows easy, fast and uncomplicated, leak-proof closure of bone defects after minimally invasive craniotomies as seen in transsphenoidal or transoral skull base approaches.

Use of scanning electron microscopy to investigate the prophylactic efficacy of rifampin-impregnated CSF shunt catheters.

Infection continues to be one of the major complications of cerebrospinal fluid (CSF) shunting procedures, and is caused mainly by skin-derived bacteria. Production of an extracellular biofilm plays an important role in the pathogenesis of shunt-associated infections by protecting bacteria from immune mechanisms and antibiotics. So far, removal of the original shunt and implantation of a new shunting device has been the only successful treatment for most patients. As an alternative strategy to prevent CSF infections, a rifampin-impregnated silicone catheter was designed to provide high initial and long-lasting (>60 days) release of bactericidal drug. To investigate the pathophysiological mechanism of its function, this new device was investigated both in vitro and in a rodent model of CSF infection by scanning electron microscopy (SEM) and bacterial culture. Staphylococcus epidermidis (10(8) cfu/ml) and S. aureus (10(4) cfu/ml) served as test strains. SEM demonstrated that, in contrast to the unloaded catheters, initial bacterial adherence on the catheter surface could be reduced to a few single cells, which did not show visible signs of proliferation. Bacterial cultures obtained simultaneously were all sterile, showing that adherent bacteria were killed immediately by the rifampin released from the catheter. Although rifampin incorporation into silicone polymers was not able to prevent initial bacterial adhesion completely, subsequent colonisation could be prevented.

Planning and simulation of neurosurgery in a virtual reality environment.

OBJECTIVE: To report our experience with preoperative neurosurgical planning in our stereoscopic virtual reality environment for 21 patients with intra- and extra-axial brain tumors and vascular malformations. METHODS: A neurosurgical planning system called VIVIAN (Virtual Intracranial Visualization and Navigation) was developed for the Dextroscope, a virtual reality environment in which the operator reaches with both hands behind a mirror into a computer-generated stereoscopic three-dimensional (3-D) object and moves and manipulates the object in real time with natural 3-D hand movements. Patient-specific data sets from multiple imaging techniques (magnetic resonance imaging, magnetic resonance angiography, magnetic resonance venography, and computed tomography) were coregistered, fused, and displayed as a stereoscopic 3-D object. A suite of 3-D tools accessible inside the VIVIAN workspace enabled users to coregister data, perform segmentation, obtain measurements, and simulate intraoperative viewpoints and the removal of bone and soft tissue. RESULTS: VIVIAN was used to plan neurosurgical procedures primarily in difficult-to-access areas, such as the cranial base and the deep brain. The intraoperative and virtual reality 3-D scenarios correlated well. The VIVIAN system substantially contributed to surgical planning by 1) providing a quick and better understanding of intracranial anatomic and abnormal spatial relationships, 2) simulating the craniotomy and the required cranial base bone work, and 3) simulating intraoperative views. CONCLUSION: The VIVIAN system allows users to work with complex imaging data in a fast, comprehensive, and intuitive manner. The 3-D interaction of this virtual reality environment is essential to the efficient assembly of surgically relevant spatial information from the data derived from multiple imaging techniques. The usefulness of the system is highly dependent on the accurate coregistration of the data and the real-time speed of the interaction.

Planning of skull base surgery in the virtual workbench: clinical experiences.

Based on the KRDL Virtual Workbench, we present a neurosurgical planning system called VIVIAN (Virtual Intracranial Visualization And Navigation). This VR environment allows for fast and intuitive interaction with three-dimensional multimodal (MRI, MRA, MRV, CT) patient specific data-sets. The user reaches behind a mirror into a 3D workspace where the 3D data is surrounded by interactive virtual tools-racks. Tumors, blood vessels, cranial nerves and surgically relevant parts of the brain are segmented by interactive control of density transfer-functions or by manual outlining and tracing tools. A neurosurgical procedure is planned by using various visualization and measurement tools and the system allows for the simulation of bone drilling and tissue removal. We have planned 16 cases which required tumor surgery at the cranial base. VIVIAN provided an efficient and comprehensive way to understand pre-operatively the complexity of anatomical and pathological relationships. The ideal craniotomy and the extent of the required skull base exposure could be specified accurately.

An interface for precise and comfortable 3D work with volumetric medical datasets.

We have developed a 3D/2D paradigm of interaction that combines manipulation of precise 3D volumetric data with unambiguous widget interaction. Precise 3D interaction is ensured by a combination of resting the lower arms on an armrest and pivoting the hands around the wrist. Unambiguous 2D interaction is achieved by providing passive haptic feedback by means of a virtual control panel whose position coincides with the physical surfaces encasing the system. We have tested this interface with a neurosurgical planning application that has been clinically used for 17 skull-base cases at two local hospitals.