Abstracts

Rationale: Considerable efforts have been made in presurgical and image-guided surgical planning towards development of novel imaging sequences and image processing algorithms to permit the integration of imaging modalities, extract pertinent anatomical information, and provide useful three-dimensional reconstructions for visualization. Neuronavigation systems often do not utilize these techniques due to the inherent technical difficulties and feasibility within clinical practice. Without these techniques, precise localization of anatomy and surgical position with these systems remains limited. In this study, we assess the accuracy of surgical localization and targeting of cortical anatomy provided by our new neuronavigation procedure developed to provides precise anatomical representation.Methods: 55 patients admitted since June 2010 at the Montreal Neurological Institute for neurosurgery were evaluated in this study. The patients underwent a variety of procedures, which included depth electrode implantation for the evaluation of epilepsy, craniotomy for epilepsy, resection of intracranial space-occupying lesions, craniotomy for arteriovenous malformation, stereotaxic biopsies and cyst aspiration. During the preoperative period, all patients underwent MRI and CTA. Patients who underwent depth electrode implantation underwent post-operative CT imaging to provide quantitative assessment of localization accuracy.Results: We developed an automated procedure which processes presurgical imaging into detailed representation of the cortical surface, cerebral vasculature, and segmented cortical gyri and sulci, for operative uses. Acquired images were post-processed to extract patient-specific anatomy, and subsequently, all data are registered and fused to permit three-dimensional volume rendering on neuronavigation systems (Medtronic StealthStation). Representation and location of cortical anatomy, cerebral vasculature, and specific cortical gyri were compared on the reconstructions yielded from this method and with our conventional navigation standard. We measured the accuracy of surgical targeting, degree of clarity (local resolution of 3D reconstruction) surgical precision (accuracy of depth electrode placements) and preparation time.Conclusions: Our procedure provided reconstructions of the cortical brain surface with precise detail, which aided in surgical localization (0.5mm) of gyral crowns, sulcal fundi, arterial and venous branches at pertinent phases, subcortical structures-of-interest, ventricles, scalp, skull, previous craniotomies, and burr holes. The degree of clarify was significantly increased (0.9mm versus 3.2mm, p < 0.005) compared with our conventional standard. This approach significantly reduced the operator time required to prepare a routine case (160 to 25 min, p < 0.001). This study demonstrates the feasibility of routine clinical use of our neuronavigation procedure for neurosurgical cases and illustrates the improvement for localization of surgical location and patient anatomy.