The 3-d reconstruction of magnetic source image data and its application in a neurosurgical guidance system

S.A.Holowka¹, S.H.Chuang¹, H. Otsubo², E. Pang², R. Sharma², A. Hunjan², J. Xiang¹, N. Chuang¹,
O. C. Snead III², J. T. Rutka³
¹Department of Diagnostic Imaging, ²Department of Neurology,³Department of Neurosurgery
The Hospital for Sick Children
555University Avenue, Toronto, Ontario, CANADA M5G 1X8

Abstract: For several years, magnetic source imaging (MSI) has been a tool utilized in functional brain mapping and for the localization of interictal spikes in patients with intractable epilepsy. This paper will discuss the production of the magnetic source images, their reconstruction into 3-Dimensional objects and the clinical application of this data in a neurosurgical guidance system.

INTRODUCTION

Magnetoencephalography data in correlation with magnetic resonance images has been utilized in determining the location of epileptic foci and eloquent brain functioning in patients with or without demonstrable lesions. On 14 patients, MEG data was collected and mapped onto MRI images. The resultant magnetic source images were converted into 3-Dimensional objects and then utilized in the operating room setting on a neurosurgical frameless stereotaxy system. This paper will discuss the data collected, the mapping of the data onto MRI, and the conversion of the MSI data into 3-Dimensional objects. It will also demonstrate the use of the MSI 2-dimensional and 3-Dimensional data sets on a Frameless Stereotaxy system.

METHODS

Magnetoencephagraphy data was collected on patients using the Omega 151 MEG system. (CTF Systems Inc., Vancouver B.C., Canada.) For 12 of the patients, simultaneous 22 lead, bipolar montage EEG data was collected simultaneously with the 151 channel MEG data. The final analysis was performed with no less than 15 data sets of 2 minutes duration with head localization being performed at the beginning and ending of each trial. Somatosensory localization was performed on each of the subjects using electrical stimulation of the median nerve. Each data-set consisted of 400 trials with the head localization being performed at the beginning and end of each collection. The somatosensory data sets were averaged, so as to improve signal noise ratio and plotted using the dipole fit program on the CTF Systems software. On some patients, auditory evoked fields and on one patient, a language localization was performed. Both of these data-sets consisted of 100 trials, with head localization being performed at the beginning and end of each collection. These data-sets were then averaged, analyzed and plotted on the dipole fit program.

After all of the MEG data was collected, a 3-D FSPGR (T1 weighted volume acquisition) MRI (GE Medical Systems, Milwaukee, Wisconsin) was acquired with fiducial markers demonstrating the position of the localization coils. The fiducials were used to co-register the exact position of the localization coils (and therefore the patient’s position in the MEG dewar) on the MRI viewer program in the CTF systems software. Analysis and plotting of all the data took place and was correlated on the MRI viewer software. The dipoles of each data set collection were marked onto the MRI images, using the CTF MARK VOXEL program in the following format:

TABLE I
Symbols for MEG dipoles

MEG DATA SET

Symbol

EEG Spikes

Filled triangle

SEF Dipoles

Filled Circle

AEF Dipoles

Filled Square

Visual Evoked Field

Hollow Shape

Language dipoles

Hollow Square

The data was transferred into the PACS system and into the ISG Allegro 3-D imaging software (Cedara Software Corp., Mississauga, Ontario). 3-Dimensional shaded surface display objects were created of the brain, the dipoles, the somatosensory cortex and the skin surface utilizing a seeded threshold method. The 2-Dimensional and 3-Dimensional data sets were then loaded from the Allegro system onto the Zeiss/SNN Neurosurgical Navigational System (Carl Zeiss Inc., North America) and this frameless stereotaxy system’s software was utilized in the operating room setting to localize directly on the patient’s brain.

RESULTS

The CTF Software program for marking dipole voxels onto the original MRI images has been the key element to this site’s use of the MSI data in the operating room. This program does not provide an overlay of MEG data onto the images but incorporates the dipole mapping into the pixel densities. In the majority of the cases, the data was used to localize epileptic foci, somatosensory cortex and motor strip. In four of the cases, the localization of the motor strip was performed and confirmed using direct brain stimulation. The application of the MSI data in the operating room setting has shown favourable results and has been able to allow the surgeon to accurately map the area of interest for subdural grid placement, or electrocorticography.

DISCUSSION

Surgical Navigation (frameless stereotaxy) has functioned as an integral tool in the practice of neurosurgery. By utilizing MSI data in this application, functional information and localization is now available to the surgeons as well as the ability to correlate both directly with the patient’s anatomy. Further investigation should be taken to confirm the accuracy of the MSI data and its utility in the operating room setting. The initial findings of this new technique have been very favourable and show great promise for continued application.

Acknowledgments: We would like to thank Tim Bardouille of CTF Systems Inc, for his assistance with the equipment. We would also like to thank Dr. Ayako Ochi and Dr. Shiro Chitoku (Division of Neurology, Hospital for Sick Children) for their assistance in the MEG program and this work.

REFERENCES

[1] H. Otsubo, R. Sharma, I. Elliott, et al.. “Confirmation of Two Magnetoencephalographic Foci by Invasive Monitoring from Subdural Electrodes in an Adolescent with Right Frontocentral Epilepsy,” Epilepsia, 40 (5):608-613, 1999.

[2] S. R. Stapleton, E. Kiriakopoulos, D. Mikulis, et al. “Combined Utility of Functional MRI, Frameless Stereotaxy and Cortical Mapping in Children with Lesions in Eloquent areas of Brain,” Pediatric Neurosurgery, 26:68-82, 1997.

[3] M. Hodaie, A. Musharbash, H. Otsubo, et. al., “Image Guided, Frameless Stereotactic Sectioning of the Corpus Callosum in Children with Intractable Epilepsy,” Pediatric Neurosurgery, 34:286-294, 2001.