Outcome based definition of the anterior thalamic deep brain stimulation target in refractory epilepsy.

BACKGROUND: Deep brain stimulation of the anterior nucleus of the thalamus (ANT) is an emerging therapy for refractory focal epilepsy. However, the most optimal target for stimulation has not been unambiguously described. OBJECTIVE: In the present study, we investigated the correlation between the stimulation site and outcome in order to define the optimal target for deep brain stimulation in refractory epilepsy. METHODS: The locations of 62 contacts used in 30 treatment attempts in 15 prospectively followed patients during a 5 year period were assessed. Treatment attempts were classified into responding and non-responding trials using seizure reduction and side effect profile as criteria. The locations of active contacts were calculated with respect to mid-commissural point and visible borders of ANT in 3T MRI (ANT-normalized coordinate system) aiming to minimize the confounding effect of individual variation in the location and size of the ANT. RESULTS: Contacts in successful treatment trials were located significantly more anterior and superior both in AC-PC and ANT-normalized coordinate systems. Favourable outcome was observed at 3T MRI based location of ANT but not at location predicted by Schaltenbrandt atlas sagittal data. Contacts used in successful trials were at anterior aspect of the ANT complex evidenced by the ANT-normalized coordinate system. CONCLUSION: The anti-epileptic effect of anterior thalamic DBS may be dependent on stimulation site especially in the anterior to posterior axis. Extensive anatomical variation confounds severely the targeting of ANT. Therefore, direct visualization of the desired target for stimulation is essential for favourable outcome in refractory epilepsy.

Defining the anterior nucleus of the thalamus (ANT) as a deep brain stimulation target in refractory epilepsy: Delineation using 3 T MRI and intraoperative microelectrode recording.

BACKGROUND: Deep brain stimulation (DBS) is a minimally invasive and reversible method to treat an increasing number of neurological and psychiatric disorders, including epilepsy. Targeting poorly defined deep structures is based in large degree on stereotactic atlas information, which may be a major source of inconsistent treatment effects. AIM OF THE STUDY: In the present study, we aimed to study whether a recently approved target for epilepsy (anterior nucleus of thalamus, ANT) is visualized in clinically established 3 T MRI and whether ANT is delineated using intraoperative microelectrode recording (MER). We have especially focused on individual variation in the location of ANT in stereotactic space. We also aimed to demonstrate the role of individual variation in interpretation of MER data by projecting samples onto AC-PC (anterior and posterior commissure) and ANT-normalized coordinate systems. METHODS: Detailed analysis of ANT delineations in 3 T MRI short tau inversion recovery (STIR) images from eight patients undergoing DBS for refractory epilepsy was performed. Coronal and sagittal cross-sectional models of ANT were plotted in the AC-PC coordinate system to study individual variation. A total of 186 MER samples collected from 10 DBS trajectories and 5 patients were analyzed, and the location of each sample was calculated and corrected accordingly to the location of the final DBS electrode and projected to the AC-PC or coordinate system normalized to ANT. RESULTS: Most of the key structures in the anatomic atlas around ANT (mammillothalamic tract and external medullary lamina) were identified in STIR images allowing visual delineation of ANT. We observed a high degree of anatomical variation in the location of ANT, and the cross-sectional areas overlapped by study patients decreased in a linear fashion with an increasing number of patients. MER information from 10 individual trajectories correlated with STIR signal characteristics by demonstrating a spike-negative zone, presumably white matter layer, at the lateral aspect of ANT in ANT-normalized coordinate system as predicted by STIR images. However, MER information projected to the AC-PC coordinate system was not able to delineate ANT. CONCLUSIONS: ANT is delineated in 3 T MRI by visualization of a thin white matter lamina between ANT and other nuclear groups that lack spiking activity. Direct targeting in the anterior thalamic area is superior to indirect targeting due to extensive individual variation in the location of ANT. Without detailed imaging information, however, a single trajectory MER has little localizing value.

Intraoperative imaging in a comprehensive neuronavigation environment for minimally invasive brain tumour surgery.

BACKGROUND: Development of an image-guided operation theatre offering multimodal information for mini-invasive neurosurgical brain tumour operations. METHODS: A multi-purpose resistive low-field MR scanner with on-off capability, was installed in a radio frequency-shielded operating room with in-room control panel and display. Intraoperative ultrasound imaging with Doppler mode as needed is used to provide check-up image data between intraoperative MR-imaging sessions. Cortical stimulation and registration are performed during awake craniotomies. The neuronavigation systems are customised arm-based and passive optical. The navigation systems show the positions of the ultrasound probe, cortical stimulation electrode, biopsy needles, endoscope and other instruments on the intraoperative MR-images. FINDINGS: Since 1999, 70 patients (mean age 47, range 3-88 years) have been operated with intraoperative MR-guidance (including 10 tumour biopsies, 56 resections). Twenty-one patients (mean age 46, range 16-67 years) underwent awake craniotomy and tumour resection secured with cortical stimulation and usually preoperative fMR-imaging. The present operating environment offered useful multimodal information for surgery of brain tumours in critical locations. Surgical mortality was 0%, morbidity included 3 (4.3%) infections and 2 (2.9%) permanent hemiparesis. Further removal of tumour was continued in 17 cases (57%) out of the 30 cases where intraoperative MR imaging was used for controlling completeness of the resection.

Versatile intraoperative MRI in neurosurgery and radiology.

BACKGROUND: Several models for the application of intra-operative magnetic resonance imaging (IMRI) have recently been reported, most of them unique. Two fundamental issues need to be addressed: optimal use of the scanner to ensure a wide base for research, development and clinical application, and an organisational model that facilitates such use. METHOD: While in our setting the IMRI project was initiated by the neurosurgeons, the need for wider use of the facilities was recognised since the beginning of the planning phase in 1996. An organisational model was developed that allowed for development of neurosurgical applications, radiological imaging, and radiological interventions and for the research and development work of the vendor. A resistive 0.23 T MR scanner was installed in a dedicated operating room environment. Unique to this scanner is the ability to turn off the magnet, allowing for normal OR activities and devices, and to turn on the magnet as needed with a relatively short six-minute ramp up time. A staged surgical technique was perfected, allowing for transfer of data to the neuronavigator outside the scanner during surgery. In neurosurgery, IMRI was used as one part of a neuronavigational system that included ultrasound imaging, intra-operative cortical stimulation during awake procedures, electrocorticography and two neuronavigators. FINDINGS: 34 neurosurgical cases included 27 brain tumour resections, 5 brain tumour biopsies, 1 extirpation of an arterio-venous malformation, and 1 haematoma evacuation. The scanner could also be used for normal clinical imaging where obese patients, children, claustophobic patients and postoperative control examinations were the major groups. The radiologists performed 110 interventions, including bone and abdominal biopsies, nerve root infiltrations and local pain therapies, with the optical needle tracking system under continuous MRI guidance. The organisational model allowed frequent use of the facilities for both neurosurgery and radiology and continuous development of the facilities. Intra-operative ultrasound was used in 20 tumour resections and in two open brain biopsies. This resulted in reduction of the number of MR imaging sessions during surgery. Five of the 27 resections were performed as awake craniotomies with cortical stimulation. For two of the resections, electrocorticography and depth electrode registrations were used. Furthermore, various non-MRI-compatible instruments and devices were used. INTERPRETATION: Intra-operative MRI is an imaging tool that can be useful especially in the context of neuronavigation. A scanner that can be turned off during surgery is particularly appropriate for neurosurgery. The concept of joint use of such facilities with other clinicians is mutually worthwhile.

Registration in interventional procedures with optical navigator.

Performing interventional procedures in the close proximity to an MR scanner widens the range of operations available for an optical tracking system. In order to gain the full benefits from both unrestricted use of surgical instruments outside the magnet and intraoperative imaging, a method for transferring the registration data of the optical navigator between two locations is required. An optical tracking system, which provides such a transfer method and tracks patient position during a surgical procedure, has been developed, tested, and demonstrated with two patient cases. J. Magn. Reson. Imaging 2001;13:93-98.