British Journal of Radiology (2003) 76, 98-103
© 2003 British Institute of Radiology
doi: 10.1259/bjr/66817309
Motor functional MRI for pre-operative and intraoperative neurosurgical guidance
I D Wilkinson, PhD, ARCP, FIPEM1,
C A J Romanowski, FRCR2,
D A Jellinek, MD, FRCS(NS)3,
J Morris, DCR1 and
P D Griffiths, FRCR, PhD1
1 Academic Unit of Radiology, University of Sheffield, Sheffield and Departments of 2 Radiology and 3 Neurosurgery, Royal Hallamshire Hospital, Sheffield Teaching Hospitals NHS Trust, Sheffield S10 2JF, UK
 |
Abstract
|
|---|
Functional MRI (fMRI) may provide a means of locating areas of eloquent cortex that can be used to guide neurosurgeons in their quest to maximize intracerebral tumour resection whilst minimizing post-procedural neurological deficits. This work aimed to develop and provide an initial assessment of such a technique. 19 patients with mass lesions close to the primary motor cortex underwent fMRI at 1.5T. A single shot echo planar technique was used to acquire data corresponding to right and left hand movement. Resultant activation maps were used to aid pre-surgical planning. Data was used in conjunction with an intraoperative navigation system in 13 cases. Activation was attributed to primary motor, primary somatosensory or supplementary motor cortex in 17 of 19 subjects. No permanent changes in motor deficit were detected post surgery. The additional information provided by fMRI, particularly when incorporated into a neuronavigation guided craniotomy, was deemed highly valuable to the neurosurgeon as it enabled safe resection of tumour in anatomical locations previously deemed to be too high risk for safe resection using conventional (non-fMRI-guided) technique. This observation is reinforced by the fact that no patients suffered permanent neurological deficit after radical tumour debulking (surgical estimates >90% tumour resection).
|
Introduction
|
|---|
Modern neurosurgical techniques aim to maximize both the survival rates of patients with resectable intracranial pathology and their quality of life following surgical intervention. The preservation of basic motor function is of great importance to the latter, particularly that of hand movement. An awareness of the spatial location of the part of the cortex responsible for affecting hand movement is important, especially when the pathology to be removed lies within its vicinity.
The coupling between performance of a motor task and regional changes in cerebral haemodynamics and oxygen saturation, monitored by blood oxygen level dependant (BOLD) functional MRI (fMRI), is well described [1, 2]. The underlying physiology has yet to be understood fully and quantified, however, there has been a drive towards the application of this coupling to the non-invasive mapping of eloquent areas of motor function [3–5]. Such maps can be used to guide surgeons in their quest to maximize tumour resection whilst minimizing motor function damage. The aim of this work was to develop and provide an initial prospective assessment of such a technique.
|
Materials and methods
|
|---|
Subjects
The subject group consisted of 19 patients (13 male, 6 female; mean age 40 years, range=19–74 years) all of whom had a known mass lesion located close to the motor cortex as demonstrated on initial CT and/or MRI. The ability to move the fingers on both hands was a pre-requisite for participation in this study. Further details of the patient group are given in Table 1
. All but two of the subjects underwent craniotomy and radical resection of the tumour, the remaining two underwent biopsy only. Ethical approval and informed consent was obtained prior to the procedure.
View this table:
[in this window]
[in a new window]
|
Table 1. Histopathological diagnosis of tumour type within, and demographic data of, the study group. Distance D1 is the minimum separation between primary motor activation and tumour boundary (defined on clinical MRI). Distance D2 is the minimum separation between primary motor activation and oedema (defined on clinical MRI). No distance is recorded where no oedema was demonstrated along a direct line between primary motor activation and tumour
|
|
Functional imaging
Imaging was performed on a standard clinical system operating at 1.5T (Eclipse; Philips Medical Systems, Cleveland, OH). A gradient recalled single shot echo planar technique (effective echo time=40 ms; repetition time=3 s; matrix=97 x 94, zero-filled to 128 x 128 over a 240 mm field of view) was used to obtain 4 mm thick transaxial images through the brain. For each functional run, one complete anatomical set of slices was acquired every 3 s for 5 min, yielding 100 time-points. A paradigm of boxcar design was used to contrast unilateral finger tapping with no finger tapping (rest). Two functional runs were performed per subject, one whilst the subject performed alternate rest (30 s) with right hand movement (30 s) and the other whilst the subject performed alternate rest with left hand movement. For each subject the transitions between rest and motor activity were caused by tactile stimulation of the lower limb. Patient compliance with the performance of the functional task was assessed by direct visual monitoring during the MR procedure.
Statistical analysis
Following on-line image reconstruction, data was transferred to an off-line workstation (UltraSPARC 30; Sun Microsystems Inc., Santa Clara, CA) for post-acquisition processing using a freely available fMRI processing and visualization package (BrainTools, http://www.aston.ac.uk/
singhkd/mri3dX). The analysis comprised three-dimensional (3D) rigid body motion correction followed by spatial smoothing to minimize any artefacts introduced by the process of rigid body motion correction (Gaussian 2 x 2 x 2 pixel full width at half maximum height (FWHM)) and the time-course of each voxel was correlated with a temporally smoothed square wave representing a model haemodynamic response [6]. This procedure utilizes a conventional correlational approach and standard parametric statistics were used for the conversion of r values to p-values. All p-values quoted were corrected for multiple-comparisons using the Bonferroni correction. The threshold that was used to define statistically significant activation was at the p![{els]](/math/els.gif)
le;0.05 (Bonferroni corrected) level.
Surgical planning/guidance
In 13 patients, areas demonstrating activation at or below the statistically significant level were manually superimposed, via qualitative radiological evaluation, on post-contrast 3D T1 weighted MRI datasets by an experienced neuroradiologist (CAJR). Following superimposition, the data were exported to a neurosurgical navigation system (BrainLAB Ltd, Surrey, UK) and used for direct intraoperative guidance. In the remaining cases, functional data superimposed by the same method upon standard qualitative imaging was used by the neurosurgeon for pre-operative surgical planning only. Areas showing activation at or below the p=0.05 level (corrected for multiple-comparisons) were not resected at surgery. A qualitative judgement was made by an experienced neurosurgeon (DAJ) at the time of surgery as to whether incorporation of the fMRI data led to the excision of less, the same or more tissue than would have been excised using standard neurosurgical technique if the fMRI data had not been available.
|
Results
|
|---|
Statistically significant correlation between voxel intensity and the model haemodynamic response was detected in 18 of 19 subjects; the remaining subject (patient # 13) did not perform the defined paradigm. One patient's functional datasets (patient # 8) were unusable due to subject movement artefact, which was apparent from the base echo planar imaging data. Functional maps demonstrating primary motor, primary somatosensory or supplementary motor activation were deemed to have been acquired in 17 of 19 (88%) subjects (Figures 1 and 2). Histopathological examination of excised tissues lead to the diagnosis of invasive intra-axial tumour in 18 of 19 patients: 11 astrocytomas (4 grade II, 4 grade III and 3 grade IV); 5 oligodendrogliomas; a secondary glioblastoma multiforme; and metastic carcinoma from a primary of unknown origin. The remaining case had an extra-axial meningioma arising from the interhemispheric fissure. In order to characterize the location of functional motor cortex with respect to the location of pathology, two distances were measured (Table 1
). D1, indicates the minimum distance between primary motor activation (on the fMRI maps) and tumour boundary (defined on standard clinical MRI), whilst D2 indicates the minimum distance between primary motor activation (on the fMRI maps) and oedema (defined on standard clinical MRI). In 8 of 17 cases, either tumour boundary or associated oedema was contiguous with activation attributed to the primary motor area. In 5 of 17 cases, the primary motor strip was not clearly defined on the underlying anatomical images on the side of the tumour but could always be defined in the contralateral hemisphere.
View larger version (100K):
[in this window]
[in a new window]
|
Figure 1. Data from a 33-year-old male with a space-occupying lesion in the right paracentral lobule depicting statistically significant correlation (p<0.05, corrected for multiple comparisons) between finger movement and model haemodynamic response overlaid (in colour) on base echo-planar images (greyscale). (a) Activation can be attributed to primary motor, primary somatosensory and supplementary motor areas of the cortex within the left hemisphere following right hand movement. (b) Bilateral activation is detected following left hand movement. (c) Post-Gd-DTPA T1 weighted spin-echo image depicting the anatomy at the level of the echo planar images (a) and (b). (d) Post-Gd-DTPA T1 weighted spin-echo image showing the tumour.
|
|
View larger version (125K):
[in this window]
[in a new window]
|
Figure 2. Data from a 32-year-old female with a space-occupying lesion within the frontal lobe depicting statistically significant correlation (p<0.05, corrected for multiple comparisons) between finger movement and model haemodynamic response overlaid (in colour) on T2 weighted fast spin echo images (greyscale). Right hand movement, (a) and (b), left (symptomatic) hand movement, (c) and (d) leading to bilateral activation. Note the proximity of activation within the right hemisphere and midline (supplementary motor area) to the tumour.
|
|
Qualitative assessment by an experienced neuroradiologist (CAJR) demonstrated greater bilateral activation with the stimulus on the symptomatic side compared with when it was on the asymptomatic side, in 13 cases (Figure 3). Qualitative assessment also demonstrated that hemispheric asymmetry in the degree of activation within the supplementary motor area was present in seven cases, more activation occurring when the symptomatic side was stimulated.
View larger version (46K):
[in this window]
[in a new window]
|
Figure 3. Data from a 30-year-old male with a right fronto-parietal lesion depicting statistically significant correlation (p<0.05, corrected for multiple comparisons) between finger movement and model haemodynamic response overlaid (in colour) on base echo-planar images (greyscale). (a) right hand movement and (b) left hand movement. Note the asymmetry in activation between hemispheres, presumably resulting from the mass-effect of the lesion.
|
|
In all cases of invasive tumour resection, the neurosurgeon estimated that more tissue was excised (>90% tumour resection) during the procedure than would have been excised using standard neurosurgical technique in the absence of intraoperative cortical mapping, if the fMRI data had not been available. Qualitative, conventional post-operative MRI (T2 and T1 weighted images) confirmed this degree of resection.
No permanent changes in neurological (motor) deficit were detected post-surgery (3 patients had transient neuro-deficit that resolved within 4 weeks of surgery).
|
Discussion
|
|---|
It is known that following radical neurosurgery, both the times to tumour progression and survival for patients with malignant intracerebral tumours are dependent upon the extent of tumour removal and the volume of any residual neoplastic cells [7, 8]. It is important that as much pathological tissue is resected as possible, however, the degree of resection must be balanced against possible risk of permanent neurological deficit and any resultant effects on post-surgical quality of life. This trade-off highlights the need for information regarding the functional anatomy associated with tasks that are fundamental to our everyday quality of life. It is reasonable to assume that the closer the pathology to areas associated with important function, the more important this functional anatomical information becomes. Such knowledge can be obtained at the time of surgery by invasive cortical mapping. However, if the location of eloquent cortex can be provided by the non-invasive technique of fMRI in advance of the surgical procedure, it would facilitate pre-operative planning. Such pre-operative planning may be further enhanced by the use of intraoperative display of the patient's functional anatomy. Intraoperative guidance based on standard 3D MRI has, in itself, been shown to be of benefit [9].
Blood oxygen-level dependent fMRI promises excellent spatial resolution. For this to be of benefit, it is important to maximize the correlation between the anatomical coordinates of "activation" derived from fMRI and the neuronal areas directly involved in the function being assessed. It has been shown that fMRI activation corresponds closely to electrophysiological data associated with epileptogenic foci [10] and with electrocortical stimulation of language, somatosensory and primary motor areas [11–15]. A comparison between fMRI and metabolic (positron emission tomography (PET)) and/or electrophysiological (transcranial magnetic stimulation, direct electrical cortical stimulation) techniques found 1 case out of 50 where fMRI and PET localization provided contradictory data [16].
Good quality "activation" maps were obtained in a high proportion of subjects within this initial study. All but three of the patients underwent surgical resection of invasive, intra-axial tumours (the other cases comprising one excision of a meningioma and two who underwent biopsy only). The pathological nature of this cohort (a high proportion of subjects had pathology le;5 mm or contiguous to fMRI-defined primary motor activation) is thought to be of importance since the radical removal of infiltrative tumour necessitates the excision of brain tissue. It is likely that such cases where the surgeon has to judge the proximity of functional anatomy to estimated tumour margin may benefit more from functional guidance than those involving non-infiltrative, extra-axial tumours. In terms of neurological outcome, our findings concur with those of Holodny et al [17], who also reported a lack of post-operative neurological deficit in a patient population of 10 (5 of whom demonstrated meningioma).
The present study found qualitative differences in the degree of bilateral activation observed following the movement of either the hand ipsilateral or the hand contralateral to the lesion. The perceived greater degree of bilateral activation was observed during movement of the hand corresponding to the symptomatic hemisphere i.e. the left hand if the tumour was in the right hemisphere. This provides evidence, albeit qualitative, for the ability of the brain to develop compensatory neuronal activation in other cortical areas. Such inference would be in agreement with Wunderlich et al [18] who reported evidence for reorganization of motor cortex to allow for preserved hand function in patients with astrocytomas.
The added information provided by fMRI in the present study, particularly when this data was incorporated into a neuronavigation-guided craniotomy, was deemed highly valuable to the neurosurgeon as it facilitated safe resection of tumour in anatomical locations previously deemed to be too high-risk for safe resection without intraoperative cortical mapping using conventional (non-fMRI-guided) technique. The strength of this argument is reinforced by the fact that no patients suffered permanent neurological deficit after radical tumour debulking (surgical estimates >90% tumour resection confirmed by post-operative MRI). In the light of this preliminary success, it would seem that extension of the technique to the mapping of functional anatomy associated with language and visual tasks [19–21] may be of direct clinical use.
The results of the present study add to the published literature which suggests that fMRI is a highly useful adjunct to standard cranial MRI when planning surgical intervention. There are many issues that may need to be resolved and warrant further study. These include possible pathology-specific effects on the measured blood-oxygen-level dependent response such as angiogenesis and vascular steal, effects due to the presence of incidental pathology such as atherosclerotic disease, and normative interindividual and intraindividual variations in cerebrovascular anatomy, compliance/autoregulation and cardiac output. There is also a plethora of possible variations in functional imaging apparatus and technique, such as the choice of imaging field strength and whether spin echo or gradient echo, echo planar imaging should be used. These possible confounding factors need to be studied if we are to determine the efficacy of fMRI intraoperative guidance accurately, as are the effects of anatomical deformations during the interventional procedure itself [22]. It must be noted that the method employed in this preliminary study relies on the radiologist's qualitative skills for the superimposition of functional and routine clinical MRI data. This is not ideal and as is the case with much radiological interpretation, open to possible observer bias. However, the results of this study provide supporting evidence for the initiation of a clinical trial comparing patients randomized to a group who have fMRI prior to an appropriate neurosurgical procedure with a group who do not.
|
Acknowledgments
|
|---|
The functional expertise of Dr D Wu of Philips Medical Systems is gratefully acknowledged. The authors wish to thank Dr K Singh for the BrainTools functional MRI analysis program.
Received for publication March 20, 2002.
Revision received September 18, 2002.
Accepted for publication December 3, 2002.
|
References
|
|---|
- Ogawa S, Lee T, Kay A, Tank D. Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proc Natl Acad Sci USA 1990;87:9868–72.[Abstract/Free Full Text]
- Boguslawska R, Romanowski CAJ, Wilkinson ID, Montaldi D, Singh KD, Walecki J. Introduction to functional magnetic resonance imaging. Med Sci Monit 1999;5:1179–86.
- Achten E, Jackson GD, Cameron JA, Abbott DF, Stella DL, Fabinyi GCA. Presurgical evaluation of the motor hand area with functional MR imaging in patients with tumors and dysplastic lesions. Radiology 1999;210:529–38.[Abstract/Free Full Text]
- Tomczak RJ, Wunderlich AP, Wang Y, et al. fMRI for preoperative neurosurgical mapping of motor cortex and language in a clinical setting. J Comput Assist Tomogr 2000;24:927–34.[CrossRef][Medline]
- Krings T, Reinges MHT, Erberich S, et al. Functional MRI for presurgical planning: problems, artefacts, and solution strategies. J Neurol Neurosurg Psychiatry 2001;70:749–60.[Abstract/Free Full Text]
- Smith AT, Greenlee MW, Singh KD, Kraemer FM, Hennig J. The processing of first- and second-order motion in human visual cortex assessed by functional magnetic resonance imaging. J Neurosci 1998;18:3816–30.[Abstract/Free Full Text]
- Keles GE, Anderson B, Berger MS. The effect of extent of resection on time to tumor progression and survival in patients with glioblastoma multiforme of the cerebral hemisphere. Surg Neurol 1999;52:371–9.[CrossRef][Medline]
- Janny P, Cure H, Mohr M, et al. Low grade supratentorial astrocytomas. Management and prognostic factors. Cancer 1994;73:1937–45.[CrossRef][Medline]
- Paleologos TS, Wadley JP, Kitchen ND, Thomas DG. Clinical utility and cost-effectiveness of interactive image-guided craniotomy: clinical comparison between conventional and image-guided meningioma surgery. Neurosurgery 2000;47:40–7.[CrossRef][Medline]
- Schwartz TH, Resor SR, De La Paz R, Goodman RR. Functional magnetic resonance imaging: Localization of ictal onset to a dysplastic cleft with simultaneous sensorimotor mapping: intraoperative electrophysiological confirmation and postoperative follow-up: Technical note. Neurosurgery 1998;43:639–45.[Medline]
- FitzGerald DB, Cosgrove GR, Ronner, et al. Location of language in the cortex: a comparison between functional MR imaging and electrocortical stimulation. AJNR 1997;18:1529–39.[Abstract]
- Lehericy S, Duffau H, Cornu P, Capelle L, Pidoux P, Carpentier A, et al. Correspondence between functional magnetic resonance imaging somatotopy of the central region: comparison with intraoperative stimulation in patients with brain tumors. J Neurosurg 2000;92:589–98.[Medline]
- Jack CR, Thompson RM, Butts RK, et al. Sensory motor cortex: correlation of presurgical mapping with functional MR imaging and invasive cortical mapping. Radiology 1994;190:85–92.[Abstract/Free Full Text]
- Fandino J, Kollias SS, Weiser HG, Valavanis A, Yonekawa Y. Intraoperative validation of functional magnetic resonance imaging and cortical reorganization patterns in patients with brain tumors involving the primary motor cortex. J Neurosurg 1999;91:238–50.[Medline]
- Puce A, Constable RT, Luby ML, McCArthy G, Nobre AC, Spencer DD, et al. Functional magnetic resonance imaging of sensory and motor cortex: comparison with electrophysiological localization. J Neurosurg 1995;83:262–70.[Medline]
- Krings T, Schreckenberger M, Rohde V, et al. Metabolic and electrophysiological validation of functional MRI. J Neurol Neurosurg Psychiatry 2001;71:762–71.[Abstract/Free Full Text]
- Holodny AI, Schulder M, Liu W-C, Wolko J, Maldjian JA, Kalnin AJ. The effect of brain tumors on BOLD functional MR imaging activation in the adjacent motor cortex: implications for image-guided neurosurgery. AJNR 2000;21:1415–22.[Abstract/Free Full Text]
- Wunderlich G, Knorr U, Herzog H, Kiwit JCW, Freund H-J, Seitz RJ. Precentral glioma location determines the displacement of cortical hand representation. Neurosurgery 1998;42:18–27.[CrossRef][Medline]
- Hirsch J, Ruge MA, Kim KHS, Correa DD, Victor JD, Relkin NR, et al. An integrated functional magnetic resonance imaging procedure for preoperative mapping of cortical areas associated with tactile, motor, language and visual functions. Neurosurgery 2000;47:711–21.[CrossRef][Medline]
- Schulder M, Holodny A, Liu W-C, Gray A, Lange G, Carmel PW. Functional magnetic resonance image-guided surgery of tumors in or near the primary visual cortex. Stereotact Funct Neurosurg 1999;73:31–6.[CrossRef][Medline]
- Roux F-E, Ibarrolla D, Lotterie J-A, Chollet F, Berry I. Perimetric visual field and functional MRI correlation: implications for image-guided surgery in occipital brain tumours. J Neurol Neurosurg Psychiatry 2001;71:505–14.[Abstract/Free Full Text]
- Hill DL, Maurer CR, Maciunas RJ, et al. Measurement of intraoperative brain surface deformation under a craniotomy. Neurosurgery 1998;43:514–26.[CrossRef][Medline]
This article has been cited by other articles:
|
|
|
|
 
M. Ragnehed, I. Hakansson, M. Nilsson, P. Lundberg, B. Soderfeldt, and M. Engstrom
Influence of Diazepam on Clinically Designed fMRI
J Neuropsychiatry Clin Neurosci,
May 1, 2007;
19(2):
164 - 172.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. R. Petrella, L. M. Shah, K. M. Harris, A. H. Friedman, T. M. George, J. H. Sampson, J. S. Pekala, and J. T. Voyvodic
Preoperative Functional MR Imaging Localization of Language and Motor Areas: Effect on Therapeutic Decision Making in Patients with Potentially Resectable Brain Tumors
Radiology,
September 1, 2006;
240(3):
793 - 802.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
B. Pirotte, C. Neugroschl, T. Metens, D. Wikler, V. Denolin, P. Voordecker, A. Joffroy, N. Massager, J. Brotchi, M. Levivier, et al.
Comparison of Functional MR Imaging Guidance to Electrical Cortical Mapping for Targeting Selective Motor Cortex Areas in Neuropathic Pain: A Study Based on Intraoperative Stereotactic Navigation
AJNR Am. J. Neuroradiol.,
October 1, 2005;
26(9):
2256 - 2266.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
K Roessler, M Donat, R Lanzenberger, K Novak, A Geissler, A Gartus, A R Tahamtan, D Milakara, T Czech, M Barth, et al.
Evaluation of preoperative high magnetic field motor functional MRI (3 Tesla) in glioma patients by navigated electrocortical stimulation and postoperative outcome
J. Neurol. Neurosurg. Psychiatry,
August 1, 2005;
76(8):
1152 - 1157.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
