¹ Surgical Planning Laboratory, MR Division, Department of Radiology, Brigham and Women's Hospital, Harvard Medical School
² Artificial Intelligence Laboratory, Massachusetts Institute of Technology
³ Division of Neurosurgery, Brigham and Women's Hospital, Harvard Medical School

Address correspondence to:

Fatma Ozlen, M.D.
Surgical Planning Laboratory, Department of Radiology, Brigham and Women's Hospital
75 Francis Street, Boston, MA 02115, USA
phone: (617) 732 7692 fax: (617) 732 7963
e-mail: fatma@bwh.harvard.edu

• Summary
• Introduction
• Methods
• Case Report
• Discussion
• Figures:
• Figure 1:The interface of the navigational system
• Figure 2:The 3D model with the position of the grid and the strip electrodes
• Figure 3:The primary motor strip and speech area
• References

Summary:

Purpose: We have developed an intraoperative optical tracking based navigational system which allows localization in the operative space. Using three-dimensional (3D) reconstruction, this system has provided precise spatial information for intraoperative cortical mapping on patients with intractable epilepsy and where the lesion lies close to the eloquent cortex.

Methods: A 23 year old male with intractable complex partial siezures presented to our institution. Proton density MR images showed a 3cm lesion which lay 2cm beneath the left frontal operculum. A 3D model of the patient was reconstructed using MR modalities. Intraoperatively, subdural grid and strips were placed over the lesion and their electrodes were registered to the 3D model which was displayed on a monitor. The navigational system was used to localize each electrode on the 3D model. By the second operation, the sites of seizure activity were established and recorded on the 3D model. A bipolar stimulator was also used determine the speech area.

Results: The lesion, which turned out to be a cortical dysplasia, was totally removed while avoiding the cortical speech area. During the postoperative period, the patient had no neurological symptoms and no seizure activity.

Conclusion: The localization of a lesion and its correlation with epileptogenic foci is important in optimizing treatment in patients with cortical dysplasia. Our navigational system was able to provide accurate localization of the lesion and its site was correlated resection with the epileptogenic foci and the related eloquent cortex. We believe that the safe surgery of the lesion has been facilitated by this system.

Key Words:

Surgical navigation - Cortical dysplasia - Epilepsy - Epilepsy surgery

Introduction:

As one of the developmental disorders of neuronal migration, cortical dysplasia is a highly epileptogenic lesion, often causing intractable partial seizures [1]. Patients with cortical dysplasia, who have medically intractable partial epilepsy, have been helped by surgical excision [1,3]. Surgical treatment of dysplasia is often difficult to assess because a significant number of these lesions, though might be seen on MR images, are not visible to the naked eye during surgery [2,3]. Furthermore, when the lesion is situated in the eloquent cortical areas, it cannot be completely removed [1,4].

We have developed a frameless, stereotactic surgical navigation system with 3D reconstruction of the brain to facilitate the localization of the seizure foci and to decrease the incidence of postoperative deficits. Preoperative planning using 3D imaging can provide visualization of the lesion and its relationship to vessels [6]. Our navigation system has allowed the accurate, real-time localization of any point in the operative space. The data obtained from cortical mapping of the margins of the lesion using this navigational system and the 3D reconstruction, increases the safety of removal of seizure foci located near the eloquent cortex.

Methods:

Three-dimensional reconstruction:

The MR data were obtained by a 1.5 Tesla MR unit (Signa; GE Medical Systems, Milwauke, WI). A series of 124 images of post-contrast, 3D SPGR (ie:Spoiled Gradient Recalled acquisition in the steady-state, 1.5 mm thickness, 256x256 matrix of 220-240 mm FOV) were obtained. An MR angiogram in the sagittal plan for 3D reconstruction of the vasculature was used. The data were digitally transferred to the Sun computer workstation (SPARC station 20; Sun Microsystems, Inc., Mountain View, CA) via a computer network.

Each image was pre-processed to reduce noise using anisotropic diffusion filtering [10]. After pre-processing, a segmentation based on signal intensities and a voxel connectivity [11, 12]) was performed using SPGR and MR angiogram. From these images, 3D objects of the skin, brain, vessels and ventricles were reconstructed using the marching cubes algorithm and a surface rendering method [11, 12, 13]. For segmentation of the lesion, proton density MR images were used because it was not apparent on the SPGR images. The segmented data of lesion was then registered to the 3D model and SPGR using the method previously described [14]. During presurgical planning, this focal lesion was considered as a possible seizure focus.

These objects were then integrated and displayed on a Sun computer workstation (Ultra-1; Sun Microsystems, Inc.) with a graphic accelerator (Creator 3D; Sun Microsystems Inc.) and 3D display software (LAVA; Sun Microsystems, Inc.). Using this program, each object may be rendered partially or completely transparent and oriented according to the viewer's choice. Each object may be individually colored to facilitate visualization.

Surgical Navigation System:

For intra-operative navigation, a real time, Light Emitting Diode-based (LED), frameless, stereotactic device developed in the laboratory was used. Prior to the beginning of surgery, a dynamic reference frame carrying three LED's (Image Guided Technologies Inc.) is fixed next to the patient's head. A series of points from the patient's skin are recorded using an LED probe (Image Guided Technologies, Inc.). The information is tracked using an optical digitizer (Flashpoint 5000; Image Guided Technologies, Inc.). The points recorded in real space are then matched to the 3D model using a two stage process. An initial rough alignment is obtained by recording the real-space location of three points, then manually matching those points on the 3D model. This initial registration is refined by finding the optimal transformation that aligns all of the points onto the skin surface of the model [15,16]; this skin-to-skin registration is acquired with high positional accuracy (< 1 mm). Following the registration the surgeon can try tracking to verify the accuracy by pointing to a known area on the patient's head. During the surgery, the surgeon uses a sterile LED probe to point to or inside the patient's brain. The surgeon is able to verify correspondence between the patient's brain and the 3D model and MRI slices by pointing to the edge of craniotomy and/or the cortical vessels. The position of all probed points may be recorded on the 3D model as a spere or arrow with interchangeable coloring. This was used to record position of the subdural electrodes.

Case Report:

History: A 23 year old male, with a history of seizures since age 2, presented to our institution. Several times a week, upon waking in the morning, he exhibited complex partial seizures with tonic features and loss of conciousness. He also had a recent onset of right hand simple partial seizures.

Presurgical examination and surgical planning: EEG revealed no definite localization of the seizure onset, although late ictal and postictal periods demonstrated bifrontal rhythmic delta activity. The proton density MR images showed a 3 cm high intensity lesion, which lay 2 cm beneath the left frontal operculum. The lesion was preoperatively thought to be "likely congenital abnormality probably a cortical dysplasia" based on MRI.

Using the technique which was mentioned above, a 3D model of the patient was reconstructed; various stuctures, including skin, brain, ventricles, vasculature and focal lesion, were segmented and then combined to created a multi-colored rotatable 3D model.

First surgery: The patient was admitted to undergo a left frontotemporal craniotomy for placement of a subdural grid and strips for the identification of the seizure focus. This system showed the location of the probe on the 3D model and the original MR images. Figure 1 shows the interface of navigation system during the actual surgery, with the location of the probe on the 3D model and MR images. This system shows the SPGR images from which the 3D model was reconstructed. Because the lesion was only detected on proton density MR images, the outline of the lesion on these images was registered to the SPGR as described previously [14].

At the surgery, there seemed to be no gross abnormalities on the patient's brain surface. A large grid was placed along the frontal lobe; two grids were placed along the temporal lobe and a single grid subtemporally and frontal inferiorly. The navigational system was used to localize the electrodes on the 3D model. Only the electrodes that were visible within the boundaries of the craniotomy area could be registered to the 3D model. Due to the rigidity of the probe, hidden electrodes could not be registered. The 3D reconstruction of those electrodes and the lesion itself was helpful in establishing lesion site and extent. The procedure was performed without complications.

Site of seizure activity was established by a week-long observation period:Before the first surgery, the EEG poorly localized the seizure activity although it suggested a left frontal origin. Using the subcortical electrodes, EEG monitoring revealed the seizure focus which coincided with the enhancing lesion. The electrodes which upon bedside stimulation resulted in spike activity, were color coded. The seizure focus was determined based on the cluster of seizure-inducing electrodes which coincided with the location of the enhancing area. This was confirmed by the registration of these electrodes to the 3D model (see Figure 2).

The second surgery: After the week-long obsevation period, the patient returned to the operating room for removal of the grids and strips and excision of the seizuri foci. Before removing the grids and strips, his speech was carefully monitored. Using the cortical stimulation (Ojemann, Radionix, Burlington, MA), Broca's area was identified and the mass which undercut this site seemed to be located anteriorly. Figure 2 showes the position of the grid and the strips plus the results of the cortical mapping displayed on the 3D model. During the second surgery and before removing the grid, the surgeon established exactly the sites of seizure activity by recording the spike-inducing electrode positions on the 3D model. The surgeon planned to attain the lesion anteriorly using navigator for guidance. The patient's speech was carefully monitored and the surgeon was able to internally decompress the lesion. Using an anterior approach, he was gradually able to remove a very large mass which undercut Broca's area without affecting the fibers which were posterior to the lesion.

This procedure was performed under local anesthesia and during this time, the patient spoke well. The lesion, cortical dysplasia, was totally removed while avoiding the adjasent speech area. Figure 3 shows the primary motor strip and speech area which were identified during the resection of the lesion.

Histopathological examination of the lesion confirmed the preoperative diagnosis.

Postoperative course: The patient did well postoperatively, without seizure activity and with a nonfocal neurologic examination.

Discussion:

Some intractable epilepsy cases which require excisional therapy cannot be treated because their epileptogenic lesions are situated at or nearby the eloquent cortical areas [4]. For such cases, multiple subpial transections (MST) has been described for the relief of the epileptogenic lesions of unresectable cortex [4,5]. It was also suggested that structural lesion consistent with the epileptiform electrical abnormality should also be removed. A directly invisible lesion such as cortical dysplasia near the eloquent cortex might have been considered for MST because of the surgical risks of total resection. We thought that if there was a chance for total resection with safety, we could prefer it.

The surgical navigator system can provide real time positional information and allow the surgeon to navigate exactly where he/she wants to be. Since a neuronavigator was reported by Watanabe et al [8], various commercial and non-commercial devices along these lines have been devised [9]. If an epileptogenic lesion is seated on the eloquent cortex, localization of the lesion is not enough for the safe and successful completion of the surgery. Because the epileptogenic focus is surrounded by functional tissue, the optimal surgical management lies in the excision of both the lesion and the epileptogenic region and preservation of functional tissue. In this point, the surgical navigator has key advantages:

1. It has been able to register multi-modal images such as phase contrast MR angiogram to MR. Visualization of the cortical surface vessels is useful for the intraoperative identification of cortical anatomy since the cortical vessels provide very important information for a surface lesion [7].
2. Preoperative planning with 3D imaging has provided selective visualization of the lesion and surrounding anatomical structures for optimal craniotomy and cortisectomy sites.
3. During the first operation, we are able to record each electrode on the 3D model.

When the second operation was performed, using a surgical navigator as a 3D localizer of each electrode, we were also able to register the data of EEG and brain mapping to the 3D model for anatomical reference and for surgical exploration. As a result, to insure the exact localization of a lesion, and the optimization of surgical treatment by combining cortical mapping and EEG results with 3D reconstruction and navigation has become possible. The navigational system allowed the real time localization within the operative space and provided spatial information for cortical mapping. The multi-modal registration is also useful to transfer functional information into 3D model for anatomical reference and for surgical exploration. We believe that these capabilities have increased the range of motion and safety of the surgery.

Figures:

Figure 1: The interface of the navigational system that is displayed on the monitor during surgery. The upper left corner displays the 3D reconstruction. The original greyscale images are also displayed. The high signal lesion is outlined (upper left and bottom).

Figure 2: The 3D model with the position of the grid and the strip electrodes after the first surgery. The lesion shown in green, the ventricles in blue and the vessels in red. The functional areas are outlined by colored circles.

Figure 3: The primary motor strip and speech area which were identified during the recession of the lesion outlined by the white triangles.

References:

1. Palmini A, Andermann F, Olivier A, et al. Focal neuronal migration disorders and intractable partial epilepsy:Results of surgical treatment. Ann Neurol 1991;30:750-757

2. Wyllie E, Comair Y.G, Kotagal P, et al. Epilepsy surgery in infants. Epilepsia 1996;37:625-637

3. Prayson R.A, Estes M.L. Cortical dysplasia: A histopathologic study of 52 cases of partial lobectomy in patients with epilepsy. Human Pathology 1995;26:493-500

4. Morrell F, Whisler W.W, Bleck T.P, et al. Multiple subpial transection: a new approach to the surgical treatment of focal epilepsy. J Neurosurg 1989;70:231-239

5. Wyler A.R, Wilkus R.J, Rostad S.W, et al. Multiple subpial transections for partial seizures in sensorimotor cortex. Neurosurg 1995;37:1122-1128

6. Kikinis R, Gleason P.L, Moriarty T.M, et al.Computer-assisted interactive three-dimensional planning for neurosurgical procedures. Neurosurg 1996;38:640-651

7. Nakajima S, Atsumi H, Kikinis R, et al. Use of cortical surface vessel registration for image-guided neurosurgery. Neurosurg 1997;40:1209

8. Watanabe E, Watanabe T, Manaka S, et al. Three dimensional digitizer (neuronavigator): new equipment for computed tomography-guided stereotaxic surgery. Surg Neurol 1987; 27:543-547

9. Barnett G.H, Kormos D.W, Steiner C.P, et al.Intraoperative localization using an armless, frameless stereotactic wand. J Neurosurg 1993;78:510-514

10. Gerig G, Kubler O, Kikinis R, et al. Nonlinear anisotropic filtering of MRI data. IEEE Trans Med Imaging 1992;11/2:221-232

11. Cline HE, Dumoulin CL, Lorensen WE, et al. 3D reconstruction of the brain from magnetic resonance imaging using a connectivity algorithm. Magn Reson Imaging 1989;5:345-352

12. Cline HE, Lorensen WE, Kikinis R, et al. Three dimensional segmentation of MR images of the head using probability and connectivity. J Comput Assist Tomogr 1990;14:1037-1045

13. Cline HE, Lorensen WE, Souza SP, et al. 3D surface rendered MR images of the brain and its vasculature. J Comput Assist Tomogr 1991;14:344-351

14. Wells, et al. Multi-modal volume registration by maximization of mutual information. Medical Image Analysis 1996;1:35-51

15. Leventon ME: A registration, tracking and visualization system for image-guided surgery. MIT Master's Thesis, May 1997

16. Grimson WEL, Ettinger GJ, White SJ, et al. An automatic registration method for frameless stereotaxy, image guided surgery, and enhanced reality visualization. IEEE Trans Med Imaging 1996;15:129-140, 1996