Journal of Medical Imaging and Radiation Oncology 58 (2014) 320–326 bs_bs_banner
RADIOLO GY —P I CTO R I A L E SSAY
Functional MRI in clinical practice: A pictorial essay Jolandi van Heerden, Patricia M Desmond and Pramit M Phal Department of Radiology, The Royal Melbourne Hospital, The University of Melbourne, Melbourne, Victoria, Australia
J van Heerden MBChB, FRANZCR; P M Desmond MSc, MDBS, FRANZCR; P M Phal MBBS, FRANZCR. Correspondence Dr Jolandi van Heerden, Department of Radiology, The Royal Melbourne Hospital, Melbourne, Victoria, Australia. Email: [email protected] Conflict of interest: None of the authors have conflicts of interest to declare.
Summary In clinical practice, functional magnetic resonance imaging (fMRI) is a valuable non-invasive tool particularly during preoperative work-up of brain tumour and epilepsy patients. In this pictorial essay, we review expected areas of eloquent cortical activation during the four major clinical paradigms, discuss pitfalls related to fMRI and look at clinical examples where fMRI was particularly valuable in preoperative planning. Key words: adult neuroimaging; applied imaging technology; magnectic resonance imaging; neuroradiology; radiography.
Disclaimer: No funding was received for this pictorial essay. All the images have been anonymised and were collected retrospectively. Submitted 20 September 2013; accepted 12 December 2013. doi:10.1111/1754-9485.12158
Background Over the past few years, functional magnetic resonance imaging (fMRI) has transitioned from the research to the clinical realm. Its great advantage lies in being able to accurately map functionally active brain regions in a noninvasive way. In the context of brain tumour surgery, preoperative knowledge of the relationship between the surgical target and adjacent eloquent cortex can assist both surgeon and patient in decision-making and planning. During a functional paradigm, neuronal stimulation triggers an increase in local cerebral metabolism and perfusion. Regional increased cerebral blood flow produces local magnetic susceptibility effects that increase MR signal via ‘blood oxygenation level dependent’ (BOLD) contrast. Imaging paradigms consist of cognitive tasks that differentially stimulate brain areas to be mapped. Typical paradigms utilised include: motor (hand, foot and face), language, memory and visual-guided saccades. The paradigm-related BOLD signal then undergoes statistical analysis to produce functional maps. 320
Functional MRI is ideally performed with a 3Tesla magnet (allowing a better signal-to-noise ratio), fitted with dedicated hardware (Fig. 1) and post-processing software operated by a well-trained neuroimaging team. In routine practice, four paradigms are tested, requiring a study duration of approximately 1 hour. Patients typically arrive 30 minutes prior to the study to facilitate training and familiarisation with the paradigms. The acquired fMRI maps are reviewed by the reporting radiologist to ensure accurate post-processing and correct statistical threshold selection before transferring the final maps to the picture archiving and communication system. Maps can be combined with anatomical images to demonstrate cortical activation with associated white matter tracts (tractography) or assist with surgical stereotaxis (Fig. 2). Local and international continued medical education courses are accessible to radiologists interested in the field, focussing on utilisation, interpretation and application of fMRI. © 2014 The Royal Australian and New Zealand College of Radiologists
Functional MRI in clinical practice
a
b
c
d
e
f
Fig. 1. Hardware and interface required to perform a functional magnetic resonance imaging (fMRI) study. (a) The fMRI synchronisation control system coordinates paradigm presentation with image acquisition, behavioural response and physiological data. (b) The projector facilitates visual presentation of the paradigm to the patient. (c and f) The scanner interface monitors behavioural and physiological data and calculates activation maps. The eye tracker is used to detect subtle patient movement that may interfere with study accuracy. (d) The patient responds to visual/auditory input by using the thumb switch. (e) The visor is mounted over the head coil to allow the patient to see the projected paradigm.
Fig. 2. Surgical stereotaxis software with eloquent cortex and associated corticospinal tracts mapped relative to the right frontal protoplasmic astrocytoma. © 2014 The Royal Australian and New Zealand College of Radiologists
321
J van Heerden et al.
a
b
c
d
Fig. 3. Variability of language laterality has been highlighted in this righthanded patient investigated for planned left hippocampectomy. Functional magnetic resonance imaging showed discordance between Broca’s and Wernicke’s area dominance – suggesting that language comprehension and expression may be affected differently by surgery. (a and c) Dominant right (thick arrow) and non-dominant left (thin arrow) Broca’s area activation confirmed on laterality index calculation. (b and d) Dominant left (thick black arrow) Wernicke’s activation in the same right-handed patient, confirmed with laterality index calculation.
Expected eloquent cortical activation during the four major clinical paradigms Language Our language task consists of sentence reading followed by a cognitive response with activation of lan-
a
322
b
c
guage and visual pathways as well as working memory. In patients with reading difficulties, the language paradigm can be performed using auditory input as well, resulting in language, visual and auditory pathway activation. Activation and laterality of Broca’s area for language expression (pars triangularis and pars opercularis) and Wernicke’s area for language comprehension (posterior aspect of the superior temporal sulcus, either superior or inferior bank) are assessed (Fig. 3).1,2 Although language dominance is usually left sided, many variations occur.1,3,4 Surgery involving the mesial temporal lobe ipsilateral to the side of language dominance can result in post-operative memory-related word finding difficulty, highlighting the value of pre-operative fMRI assessment.5 Also expected is activation of an extensive visual and cognitive network that facilitates synchronised eye movements as well as an understanding of visual images and their spatial relationships.1 These areas include the frontal eye fields (precentral sulcus), supplementary eye fields (medial frontal cortex), visuospatial attention region (intraparietal sulcus), visuospatial processing region (temporoparietal junction), visual motion detection (middle temporal cortex/V5), visual cortex (occipital lobes – V1 to V4) and executive working memory (dorsolateral prefrontal cortex) (Fig. 4).1,2
Memory Three different forms of memory are considered – episodic memory, working memory and semantic memory. In the clinical setting, mainly episodic memory is tested (encoding and retrieval), with activation of the hippocampi (memory), dorsolateral prefrontal cortex (working memory, executive function) and anterior cingulate gyrus (attention) (Fig. 5).1 Working memory forms background awareness/ decision-making while performing other tasks, and semantic memory relates to ideas and knowledge about the world.1,6 Fig. 4. Although the language paradigm is primarily assessed, additional cognitive and visual activation is also demonstrated. (a) Activation of bilateral Wernicke’s areas for language comprehension (thick blue arrows) as well as the primary visual fields V1-V4 (thin black arrows) and visual motion detection V5 (thick black arrow). (b) Broca’s area activation for language expression on the right (thick blue arrow) as well as V1–V4 (thin black arrows) and V5 (thick black arrow) activation. (c) Frontal eye field activation (thick black arrows), supplementary eye field activation (yellow arrow) and visuospatial attention (thin black arrows).
© 2014 The Royal Australian and New Zealand College of Radiologists
Functional MRI in clinical practice
Fig. 5. Memory paradigm activation. (a–c) Activation of bilateral hippocampi (black arrows) as well as the mesial temporal lobes (blue arrows), responsible for working memory. Although it is often the least robust of the paradigms, it can provide additional information regarding the possibility of memory impairment following mesial temporal surgery.
a
b
a
b
c
c
d
Fig. 6. Visual field activation. (a) Activation is noted within the primary visual cortex V1–V4 (black arrows) as well as the motion detection cortex V5/middle temporal cortex (red arrows). (b) Visuospatial attention (black arrows) and visuospatial processing (red arrow) activation. (c) Activation of bilateral frontal eye fields (long white arrows) as well as the supplementary eye fields (yellow arrow) with right dorsolateral prefrontal cortex activation (short thick white arrow) related to working memory. (d) Activation of the right lateral geniculate nucleus (thick black arrow) and the primary visual cortex V1–V4 (thin black arrows).
Vision Visual-guided saccades are tested by following a moving dot. This results in activation of the frontal eye fields (precentral sulcus), supplementary eye fields (medial frontal cortex), visual cortex (occipital lobe V1–V4), visual motion detection region (middle temporal cortex/V5), visuospatial attention (intra-parietal sulcus), visuospatial processing (temporoparietal junction) and lateral geniculate nuclei (part of visual pathway) (Fig. 6).1,7
a
b
Motor Motor tasks are typically performed using hand, foot or tongue/mouth movement signalled via a visual cue. The areas of activation include the respective regions on the homunculus responsible for movement in the primary motor cortex (pre-central gyrus), supplementary motor cortex (medial frontal lobe), frontal eye field (pre-central sulcus), supplementary eye field (medial frontal lobe anterior to the supplementary motor area) and motor coordination (thalamus and cerebellum) (Fig. 7).1,8,9 © 2014 The Royal Australian and New Zealand College of Radiologists
Fig. 7. Motor paradigm activation. (a) Activation of bilateral hand primary motor and sensory cortices (long white arrows) as well as activation of the supplementary motor cortex (short white arrow). (b) Bilateral activation of the tongue/mouth primary motor cortex (thick black arrows) and primary left sensory cortex (yellow arrow). There is also activation of the left thalamus (blue arrow) and bilateral basal ganglia (short white arrows) for motor integration and coordination.
323
J van Heerden et al.
b
a
c
Fig. 8. Functional MRI maps obtained as part of the preoperative work-up of a left frontal protoplasmic astrocytoma (WHOII). The clinical concern in this case was the relationship of the left-sided Broca’s area and the left-sided motor cortex with the tumour. It was shown that the tumour bordered on the left-sided activation areas of both tongue movement and language expression and the functional magnetic resonance imaging study alerted the neurosurgical team to the fact that the left Broca’s area was inferiorly displaced relative to its usual expected area. (a) Bilateral motor tongue activation was demonstrated; on the left, the tongue motor cortex contacted the posterior tumour margin (black arrow). (b) The dominant left Broca’s area was displaced inferiorly by the tumour (thick black arrow) compared with the expected location noted on the right (thin black arrow). (c) Left-sided hand motor activation (arrow) was confirmed to be well away from tumour.
b
a
Fig. 9. Motor paradigms for hand and foot were performed to assess the motor strip in relation to this right fronto-parietal diffuse astrocytoma (WHOII). The functional magnetic resonance imaging assisted the neurosurgical team by demonstrating an unusually anteriorly displaced right-sided foot motor activation region. (a) Bilateral primary hand motor activation was demonstrated (thick black arrows) – right-sided activation contacts the lateral margin of the tumour. The supplementary motor area activation (white arrow) was seen antero-medial to the lesion. (b) With the supplementary motor area identified during hand motion, new mesial frontal activation seen during foot motion confirmed an abnormally anteriorly displaced right hemispheric primary foot motor cortex (yellow arrow) relative to its normal position on the left (white arrow).
a
b
a
b
Fig. 11. Language paradigm testing during preoperative tumour work-up. The clinical question in these two cases was the location of the dominant left-sided language centres in relation to the tumours. (a) Left fronto-temporal diffuse astrocytoma (WHOII). Functional magnetic resonance imaging (fMRI) demonstrated tumour involvement of the left dominant Broca’s area (arrow). (b) A different patient with a left parieto-temporal glioblastoma multiforme – fMRI confirmed that the dominant left Wernicke’s area (black arrow) was well anterior to the tumour margin (red arrow), suggesting that surgical resection could be achieved.
c
Fig. 10. Language paradigm testing in a right-handed patient with a left temporal ganglioglioma. The clinical issues in this case were language dominance and the relationship between the tumour and the left-sided Broca’s and Wernicke’s areas. (a) As expected in a right-handed patient, the right-sided Broca’s (black arrow) and Wernicke’s (white arrow) areas were non-dominant, highlighting the risk of receptive/expressive dysphasia should the tumour resection encroach on the dominant left language centres. (b and c) Localisation of the dominant left-sided language centres demonstrated an antero-superiorly displaced Broca’s area (thick black arrow) and a postero-superiorly displaced left Wernicke’s area (thick white arrow) as a result of mass effect.
324
© 2014 The Royal Australian and New Zealand College of Radiologists
Functional MRI in clinical practice
Fig. 12. (a) Right temporal encephalomalacia secondary to prior herpes encephalitis causing treatment refractory seizures. The patient was investigated for a planned right temporal lobectomy and right hippocampectomy. The clinical question related to language dominance to assist with surgical planning and to determine the potential for post-procedural memory-related word finding/comprehension difficulties. (b and c) Preoperative fMRI demonstrated co-dominant Broca’s areas (as shown in b) and left-dominant Wernicke’s area, suggesting that language would be less severely (as shown in c) affected following surgery.
a
b
a
b
c
c
d
Fig. 13. (a) Right amygdala dysplasia (arrow) resulting in treatment-refractory epilepsy. The clinical question was language dominance. (b–d) Despite the patient being right handed, right-sided Broca’s dominance was demonstrated (as seen in b). Wernicke’s co-dominance was shown with functional maps (shown in c) and confirmed with laterality index testing (shown in d).
fMRI shortcomings and pitfalls
Conclusion
The success of a fMRI study is heavily reliant on patient compliance. Patient movement, language barriers, impaired task compliance and cognitive impairment can greatly limit study efficacy. Other factors that can interfere with cortical activation include steno-occlusive disease involving the arteries of the head or neck, vascular steal in arterio-venous malformations and magnetic field inhomogeneities (including blood product deposition and surgical clips) interfering with the BOLD signal.1
Functional MRI can play an important part in preoperative planning and assessment. The reporting radiologist needs to have a thorough knowledge of expected cortical activation during the four major clinical paradigms and be aware of potential imaging pitfalls that may affect the quality and accuracy of the investigation.
Preoperative imaging examples Pathological processes can involve or displace eloquent cortical areas. Functional MRI can assist with noninvasive preoperative cortical mapping. This provides additional information for pre-surgical discussion and decision-making, as well as avoiding the need for awake craniotomy with intraoperative cortical stimulation (Figs 8–13).10–12 © 2014 The Royal Australian and New Zealand College of Radiologists
References: 1. Faro SH, Mohamed FB, Law M. Functional Neuroradiology Principles and Clinical Applications. Springer, New York, 2011; 350–507. 2. Bookheimer S. Pre-surgical language mapping with functional magnetic resonance imaging. Neuropsychol Rev 2007; 17: 145–55. 3. Khedr EM, Hamed E, Said A, Basahi J. Handedness and language cerebral lateralization. Eur J Appl Physiol 2002; 87: 469–73. 4. Szaflarski JP, Holland SK, Schmithorst VJ, Byars AJ. An fMRI study of language lateralization in children and adults. Hum Brain Mapp 2006; 27: 202–12.
325
J van Heerden et al.
5. Rabin ML, Narayan VM, Kimberg DY, Casasantol DJ, Glosser G, Tracy JI et al. Functional MRI predicts post-surgical memory following temporal lobectomy. Brain 2004; 127: 2286–98. 6. Thompson-Schill SL. Neuroimaging studies of semantic memory: inferring how from where. Neuropsychol 2003; 41: 280–92. 7. Li W, Wait S, Ogg RJ, Scoggins MA, Zou P, Wheless J, Boop FA. Functional magnetic resonance imaging of the visual cortex performed in children under sedation to assist in presurgical planning. J Neurosurg 2013; 11: 543–6. 8. Bizzi A, Blasi V, Falini A, Ferroli P, Cadioli M et al. Presurgical functional MR imaging of language and motor functions: validation with intraoperative electrocortical mapping. Radiology 2008; 248: 579–89.
326
9. Nair DG, Purcott KL, Fuchs A, Steinberg F, Kelso JAS. Cortical and cerebellar activity of the human brain during imagined and executed unimanual and bimanual action sequences: a functional MRI study. Cogn Brain Res 2003; 15: 250–60. 10. Orringer DA, Vago DR, Golby AJ. Clinical applications and future directions of functional MRI. Semin Neurol 2012; 32: 466–75. 11. Gumprecht H, Ebel GK, Auer DP, Lumenta CB. Neuronavigation and functional MRI for surgery in patients with lesions in eloquent brain areas. Minim Invasive Neurosurg 2002; 45: 151–3. 12. Pillai JJ. The evolution of clinical functional imaging during the past two decades and its current impact on neurosurgical planning. Am J Neuroradiol 2010; 31: 219–25.
© 2014 The Royal Australian and New Zealand College of Radiologists
RADIOLO GY —P I CTO R I A L E SSAY
Functional MRI in clinical practice: A pictorial essay Jolandi van Heerden, Patricia M Desmond and Pramit M Phal Department of Radiology, The Royal Melbourne Hospital, The University of Melbourne, Melbourne, Victoria, Australia
J van Heerden MBChB, FRANZCR; P M Desmond MSc, MDBS, FRANZCR; P M Phal MBBS, FRANZCR. Correspondence Dr Jolandi van Heerden, Department of Radiology, The Royal Melbourne Hospital, Melbourne, Victoria, Australia. Email: [email protected] Conflict of interest: None of the authors have conflicts of interest to declare.
Summary In clinical practice, functional magnetic resonance imaging (fMRI) is a valuable non-invasive tool particularly during preoperative work-up of brain tumour and epilepsy patients. In this pictorial essay, we review expected areas of eloquent cortical activation during the four major clinical paradigms, discuss pitfalls related to fMRI and look at clinical examples where fMRI was particularly valuable in preoperative planning. Key words: adult neuroimaging; applied imaging technology; magnectic resonance imaging; neuroradiology; radiography.
Disclaimer: No funding was received for this pictorial essay. All the images have been anonymised and were collected retrospectively. Submitted 20 September 2013; accepted 12 December 2013. doi:10.1111/1754-9485.12158
Background Over the past few years, functional magnetic resonance imaging (fMRI) has transitioned from the research to the clinical realm. Its great advantage lies in being able to accurately map functionally active brain regions in a noninvasive way. In the context of brain tumour surgery, preoperative knowledge of the relationship between the surgical target and adjacent eloquent cortex can assist both surgeon and patient in decision-making and planning. During a functional paradigm, neuronal stimulation triggers an increase in local cerebral metabolism and perfusion. Regional increased cerebral blood flow produces local magnetic susceptibility effects that increase MR signal via ‘blood oxygenation level dependent’ (BOLD) contrast. Imaging paradigms consist of cognitive tasks that differentially stimulate brain areas to be mapped. Typical paradigms utilised include: motor (hand, foot and face), language, memory and visual-guided saccades. The paradigm-related BOLD signal then undergoes statistical analysis to produce functional maps. 320
Functional MRI is ideally performed with a 3Tesla magnet (allowing a better signal-to-noise ratio), fitted with dedicated hardware (Fig. 1) and post-processing software operated by a well-trained neuroimaging team. In routine practice, four paradigms are tested, requiring a study duration of approximately 1 hour. Patients typically arrive 30 minutes prior to the study to facilitate training and familiarisation with the paradigms. The acquired fMRI maps are reviewed by the reporting radiologist to ensure accurate post-processing and correct statistical threshold selection before transferring the final maps to the picture archiving and communication system. Maps can be combined with anatomical images to demonstrate cortical activation with associated white matter tracts (tractography) or assist with surgical stereotaxis (Fig. 2). Local and international continued medical education courses are accessible to radiologists interested in the field, focussing on utilisation, interpretation and application of fMRI. © 2014 The Royal Australian and New Zealand College of Radiologists
Functional MRI in clinical practice
a
b
c
d
e
f
Fig. 1. Hardware and interface required to perform a functional magnetic resonance imaging (fMRI) study. (a) The fMRI synchronisation control system coordinates paradigm presentation with image acquisition, behavioural response and physiological data. (b) The projector facilitates visual presentation of the paradigm to the patient. (c and f) The scanner interface monitors behavioural and physiological data and calculates activation maps. The eye tracker is used to detect subtle patient movement that may interfere with study accuracy. (d) The patient responds to visual/auditory input by using the thumb switch. (e) The visor is mounted over the head coil to allow the patient to see the projected paradigm.
Fig. 2. Surgical stereotaxis software with eloquent cortex and associated corticospinal tracts mapped relative to the right frontal protoplasmic astrocytoma. © 2014 The Royal Australian and New Zealand College of Radiologists
321
J van Heerden et al.
a
b
c
d
Fig. 3. Variability of language laterality has been highlighted in this righthanded patient investigated for planned left hippocampectomy. Functional magnetic resonance imaging showed discordance between Broca’s and Wernicke’s area dominance – suggesting that language comprehension and expression may be affected differently by surgery. (a and c) Dominant right (thick arrow) and non-dominant left (thin arrow) Broca’s area activation confirmed on laterality index calculation. (b and d) Dominant left (thick black arrow) Wernicke’s activation in the same right-handed patient, confirmed with laterality index calculation.
Expected eloquent cortical activation during the four major clinical paradigms Language Our language task consists of sentence reading followed by a cognitive response with activation of lan-
a
322
b
c
guage and visual pathways as well as working memory. In patients with reading difficulties, the language paradigm can be performed using auditory input as well, resulting in language, visual and auditory pathway activation. Activation and laterality of Broca’s area for language expression (pars triangularis and pars opercularis) and Wernicke’s area for language comprehension (posterior aspect of the superior temporal sulcus, either superior or inferior bank) are assessed (Fig. 3).1,2 Although language dominance is usually left sided, many variations occur.1,3,4 Surgery involving the mesial temporal lobe ipsilateral to the side of language dominance can result in post-operative memory-related word finding difficulty, highlighting the value of pre-operative fMRI assessment.5 Also expected is activation of an extensive visual and cognitive network that facilitates synchronised eye movements as well as an understanding of visual images and their spatial relationships.1 These areas include the frontal eye fields (precentral sulcus), supplementary eye fields (medial frontal cortex), visuospatial attention region (intraparietal sulcus), visuospatial processing region (temporoparietal junction), visual motion detection (middle temporal cortex/V5), visual cortex (occipital lobes – V1 to V4) and executive working memory (dorsolateral prefrontal cortex) (Fig. 4).1,2
Memory Three different forms of memory are considered – episodic memory, working memory and semantic memory. In the clinical setting, mainly episodic memory is tested (encoding and retrieval), with activation of the hippocampi (memory), dorsolateral prefrontal cortex (working memory, executive function) and anterior cingulate gyrus (attention) (Fig. 5).1 Working memory forms background awareness/ decision-making while performing other tasks, and semantic memory relates to ideas and knowledge about the world.1,6 Fig. 4. Although the language paradigm is primarily assessed, additional cognitive and visual activation is also demonstrated. (a) Activation of bilateral Wernicke’s areas for language comprehension (thick blue arrows) as well as the primary visual fields V1-V4 (thin black arrows) and visual motion detection V5 (thick black arrow). (b) Broca’s area activation for language expression on the right (thick blue arrow) as well as V1–V4 (thin black arrows) and V5 (thick black arrow) activation. (c) Frontal eye field activation (thick black arrows), supplementary eye field activation (yellow arrow) and visuospatial attention (thin black arrows).
© 2014 The Royal Australian and New Zealand College of Radiologists
Functional MRI in clinical practice
Fig. 5. Memory paradigm activation. (a–c) Activation of bilateral hippocampi (black arrows) as well as the mesial temporal lobes (blue arrows), responsible for working memory. Although it is often the least robust of the paradigms, it can provide additional information regarding the possibility of memory impairment following mesial temporal surgery.
a
b
a
b
c
c
d
Fig. 6. Visual field activation. (a) Activation is noted within the primary visual cortex V1–V4 (black arrows) as well as the motion detection cortex V5/middle temporal cortex (red arrows). (b) Visuospatial attention (black arrows) and visuospatial processing (red arrow) activation. (c) Activation of bilateral frontal eye fields (long white arrows) as well as the supplementary eye fields (yellow arrow) with right dorsolateral prefrontal cortex activation (short thick white arrow) related to working memory. (d) Activation of the right lateral geniculate nucleus (thick black arrow) and the primary visual cortex V1–V4 (thin black arrows).
Vision Visual-guided saccades are tested by following a moving dot. This results in activation of the frontal eye fields (precentral sulcus), supplementary eye fields (medial frontal cortex), visual cortex (occipital lobe V1–V4), visual motion detection region (middle temporal cortex/V5), visuospatial attention (intra-parietal sulcus), visuospatial processing (temporoparietal junction) and lateral geniculate nuclei (part of visual pathway) (Fig. 6).1,7
a
b
Motor Motor tasks are typically performed using hand, foot or tongue/mouth movement signalled via a visual cue. The areas of activation include the respective regions on the homunculus responsible for movement in the primary motor cortex (pre-central gyrus), supplementary motor cortex (medial frontal lobe), frontal eye field (pre-central sulcus), supplementary eye field (medial frontal lobe anterior to the supplementary motor area) and motor coordination (thalamus and cerebellum) (Fig. 7).1,8,9 © 2014 The Royal Australian and New Zealand College of Radiologists
Fig. 7. Motor paradigm activation. (a) Activation of bilateral hand primary motor and sensory cortices (long white arrows) as well as activation of the supplementary motor cortex (short white arrow). (b) Bilateral activation of the tongue/mouth primary motor cortex (thick black arrows) and primary left sensory cortex (yellow arrow). There is also activation of the left thalamus (blue arrow) and bilateral basal ganglia (short white arrows) for motor integration and coordination.
323
J van Heerden et al.
b
a
c
Fig. 8. Functional MRI maps obtained as part of the preoperative work-up of a left frontal protoplasmic astrocytoma (WHOII). The clinical concern in this case was the relationship of the left-sided Broca’s area and the left-sided motor cortex with the tumour. It was shown that the tumour bordered on the left-sided activation areas of both tongue movement and language expression and the functional magnetic resonance imaging study alerted the neurosurgical team to the fact that the left Broca’s area was inferiorly displaced relative to its usual expected area. (a) Bilateral motor tongue activation was demonstrated; on the left, the tongue motor cortex contacted the posterior tumour margin (black arrow). (b) The dominant left Broca’s area was displaced inferiorly by the tumour (thick black arrow) compared with the expected location noted on the right (thin black arrow). (c) Left-sided hand motor activation (arrow) was confirmed to be well away from tumour.
b
a
Fig. 9. Motor paradigms for hand and foot were performed to assess the motor strip in relation to this right fronto-parietal diffuse astrocytoma (WHOII). The functional magnetic resonance imaging assisted the neurosurgical team by demonstrating an unusually anteriorly displaced right-sided foot motor activation region. (a) Bilateral primary hand motor activation was demonstrated (thick black arrows) – right-sided activation contacts the lateral margin of the tumour. The supplementary motor area activation (white arrow) was seen antero-medial to the lesion. (b) With the supplementary motor area identified during hand motion, new mesial frontal activation seen during foot motion confirmed an abnormally anteriorly displaced right hemispheric primary foot motor cortex (yellow arrow) relative to its normal position on the left (white arrow).
a
b
a
b
Fig. 11. Language paradigm testing during preoperative tumour work-up. The clinical question in these two cases was the location of the dominant left-sided language centres in relation to the tumours. (a) Left fronto-temporal diffuse astrocytoma (WHOII). Functional magnetic resonance imaging (fMRI) demonstrated tumour involvement of the left dominant Broca’s area (arrow). (b) A different patient with a left parieto-temporal glioblastoma multiforme – fMRI confirmed that the dominant left Wernicke’s area (black arrow) was well anterior to the tumour margin (red arrow), suggesting that surgical resection could be achieved.
c
Fig. 10. Language paradigm testing in a right-handed patient with a left temporal ganglioglioma. The clinical issues in this case were language dominance and the relationship between the tumour and the left-sided Broca’s and Wernicke’s areas. (a) As expected in a right-handed patient, the right-sided Broca’s (black arrow) and Wernicke’s (white arrow) areas were non-dominant, highlighting the risk of receptive/expressive dysphasia should the tumour resection encroach on the dominant left language centres. (b and c) Localisation of the dominant left-sided language centres demonstrated an antero-superiorly displaced Broca’s area (thick black arrow) and a postero-superiorly displaced left Wernicke’s area (thick white arrow) as a result of mass effect.
324
© 2014 The Royal Australian and New Zealand College of Radiologists
Functional MRI in clinical practice
Fig. 12. (a) Right temporal encephalomalacia secondary to prior herpes encephalitis causing treatment refractory seizures. The patient was investigated for a planned right temporal lobectomy and right hippocampectomy. The clinical question related to language dominance to assist with surgical planning and to determine the potential for post-procedural memory-related word finding/comprehension difficulties. (b and c) Preoperative fMRI demonstrated co-dominant Broca’s areas (as shown in b) and left-dominant Wernicke’s area, suggesting that language would be less severely (as shown in c) affected following surgery.
a
b
a
b
c
c
d
Fig. 13. (a) Right amygdala dysplasia (arrow) resulting in treatment-refractory epilepsy. The clinical question was language dominance. (b–d) Despite the patient being right handed, right-sided Broca’s dominance was demonstrated (as seen in b). Wernicke’s co-dominance was shown with functional maps (shown in c) and confirmed with laterality index testing (shown in d).
fMRI shortcomings and pitfalls
Conclusion
The success of a fMRI study is heavily reliant on patient compliance. Patient movement, language barriers, impaired task compliance and cognitive impairment can greatly limit study efficacy. Other factors that can interfere with cortical activation include steno-occlusive disease involving the arteries of the head or neck, vascular steal in arterio-venous malformations and magnetic field inhomogeneities (including blood product deposition and surgical clips) interfering with the BOLD signal.1
Functional MRI can play an important part in preoperative planning and assessment. The reporting radiologist needs to have a thorough knowledge of expected cortical activation during the four major clinical paradigms and be aware of potential imaging pitfalls that may affect the quality and accuracy of the investigation.
Preoperative imaging examples Pathological processes can involve or displace eloquent cortical areas. Functional MRI can assist with noninvasive preoperative cortical mapping. This provides additional information for pre-surgical discussion and decision-making, as well as avoiding the need for awake craniotomy with intraoperative cortical stimulation (Figs 8–13).10–12 © 2014 The Royal Australian and New Zealand College of Radiologists
References: 1. Faro SH, Mohamed FB, Law M. Functional Neuroradiology Principles and Clinical Applications. Springer, New York, 2011; 350–507. 2. Bookheimer S. Pre-surgical language mapping with functional magnetic resonance imaging. Neuropsychol Rev 2007; 17: 145–55. 3. Khedr EM, Hamed E, Said A, Basahi J. Handedness and language cerebral lateralization. Eur J Appl Physiol 2002; 87: 469–73. 4. Szaflarski JP, Holland SK, Schmithorst VJ, Byars AJ. An fMRI study of language lateralization in children and adults. Hum Brain Mapp 2006; 27: 202–12.
325
J van Heerden et al.
5. Rabin ML, Narayan VM, Kimberg DY, Casasantol DJ, Glosser G, Tracy JI et al. Functional MRI predicts post-surgical memory following temporal lobectomy. Brain 2004; 127: 2286–98. 6. Thompson-Schill SL. Neuroimaging studies of semantic memory: inferring how from where. Neuropsychol 2003; 41: 280–92. 7. Li W, Wait S, Ogg RJ, Scoggins MA, Zou P, Wheless J, Boop FA. Functional magnetic resonance imaging of the visual cortex performed in children under sedation to assist in presurgical planning. J Neurosurg 2013; 11: 543–6. 8. Bizzi A, Blasi V, Falini A, Ferroli P, Cadioli M et al. Presurgical functional MR imaging of language and motor functions: validation with intraoperative electrocortical mapping. Radiology 2008; 248: 579–89.
326
9. Nair DG, Purcott KL, Fuchs A, Steinberg F, Kelso JAS. Cortical and cerebellar activity of the human brain during imagined and executed unimanual and bimanual action sequences: a functional MRI study. Cogn Brain Res 2003; 15: 250–60. 10. Orringer DA, Vago DR, Golby AJ. Clinical applications and future directions of functional MRI. Semin Neurol 2012; 32: 466–75. 11. Gumprecht H, Ebel GK, Auer DP, Lumenta CB. Neuronavigation and functional MRI for surgery in patients with lesions in eloquent brain areas. Minim Invasive Neurosurg 2002; 45: 151–3. 12. Pillai JJ. The evolution of clinical functional imaging during the past two decades and its current impact on neurosurgical planning. Am J Neuroradiol 2010; 31: 219–25.
© 2014 The Royal Australian and New Zealand College of Radiologists
