The expected and unexpected Initial experiences in a

The expected and unexpected: Initial experiences in a de novo f. MRI program Tannenbaum AD, Sakai O, Jara H, Barest G, Norbash AM, Mian AZ Department of Radiology Boston University School of Medicine Boston Medical Center Boston, MA

Introduction Basic Principles Normal Anatomy Paradigms Unusual Patterns Discussion Functional magnetic resonance imaging (f. MRI) of the _ brain allows for identification of eloquent areas for preoperative planning and for various psychiatric and neurological disorders. -Predominantly used for preoperative planning -Localization is achieved using various paradigms designed to stimulate specific areas such as motor, speech, and sensory regions -Precise preoperative localization enables maximal resection and may also help predict postoperative deficits -High concordance rates with Wada testing and intraoperative cortical stimulation -Has been shown to alter surgical management -Data is generated using paradigms and BOLD imaging Mahvash, M. , Maslehaty, H. , Jansen, O. , Mehdorn, H. M. , & Petridis, A. K. (2014). Functional magnetic resonance imaging of motor and language for preoperative planning of neurosurgical procedures adjacent to functional areas. Clinical Neurology and Neurosurgery, 123, 72– 77. doi: 10. 1016/j. clineuro. 2014. 05. 011 Medina, L. S. , Aguirre, E. , Bernal, B. , & Altman, N. R. (2004). Functional MR imaging versus Wada test for evaluation of language lateralization: cost analysis. Radiology, 230(1), 49– 54. doi: 10. 1148/radiol. 2301021122 Petrella, J. R. , Shah, L. M. , Harris, K. M. , Friedman, A. H. , George, T. M. , Sampson, J. H. , et al. (2006). Preoperative functional MR imaging localization of language and motor areas: effect on therapeutic decision making in patients with potentially resectable brain tumors. Radiology, 240(3), 793– 802. doi: 10. 1148/radiol. 2403051153

Introduction Basic Principles Normal Anatomy BOLD Hardware/Software Paradigms Unusual Patterns Discussion _ Blood Oxygen Level-dependent Imaging (BOLD) Basic Principles of BOLD -Originally discovered in 1990 by Ogawa et al -Deoxyhemoglobin (deoxy. Hb) acts as an endogenous paramagnetic contrast agent for BOLD -Images are acquired using echoplanar imaging -Cortical activation leads to a drop in oxy. HB and increase in CO 2 with a compensatory increase in blood flow, leading to higher oxy. Hb and dilution of deoxy. Hb -BOLD relies on the gradient of oxy. Hb: deoxy. Hb to generate functional data Kim, S. -G. , & Ogawa, S. (2012). Biophysical and physiological origins of blood oxygenation level-dependent f. MRI signals. Journal of Cerebral Blood Flow and Metabolism : Official Journal of the International Society of Cerebral Blood Flow and Metabolism, 32(7), 1188– 1206. doi: 10. 1038/jcbfm. 2012. 23

Introduction _ Basic Principles Normal Anatomy BOLD Hardware/Software f. MRI Paradigms Unusual Patterns Discussion Stimulus Short T 2*: Long T 2*: deoxygenated MRI Pulse Sequence Stimulus-dependent MRI signal change Neurovascular Coupling Hemodynamic Response

Main BOLD response Fig. 1: Schematic showing the time course of BOLD response to a short stimulus. The fast response has a negative peak at about two seconds poststimulus due to brief decrease in blood oxygenation after neural activity. The main BOLD response peaks at about five seconds with FWHM(full width at half maximum) of about four seconds. The signal takes about a minute to return to baseline.

Introduction Basic Principles Normal Anatomy BOLD Hardware/Software _Schematic of BOLD Paradigms Unusual Patterns Discussion Increased oxy. Hb signal generation Increased blood flow Cortical Stimulation Increased Neuronal Activity Decreased deoxy. Hb concentration Increased O 2 Extraction Increased deoxy. Hb f. MRI signal deoxy. Hb has a paramagnetic effect; Increased blood flow is sufficient to dilute deoxy. Hb concentration

Introduction Basic Principles Normal Anatomy BOLD Hardware/Software Paradigms Unusual Patterns Discussion Blood Oxygen Level-dependent Imaging (BOLD) Resting state schematic illustrating the paramagnetic effect of deoxy. Hb Hb. O 2 Hb Hb Distorted field lines Activated state with increase in oxy. Hb. Measurement of the gradient between oxy. Hb and deoxy. Hb will yield BOLD signal. Hb. O 2 Adapted from Daniel Marcus Ph. D. , “Brain Imaging for fun and Profit” Hb. O 2 Hb

Introduction Basic Principles Normal Anatomy BOLD Hardware/Software Paradigms Unusual Patterns Discussion _ Blood Oxygen Level-dependent Imaging (BOLD) Limitations of BOLD -Relies on T 2* imaging which is sensitive to magnetic field inhomogeneity -Very sensitive to patient motion -An indirect measurement of cortical neuronal activity. Data reflects changes in blood flow rather than direct neuronal activity Kim, S. -G. , & Ogawa, S. (2012). Biophysical and physiological origins of blood oxygenation level-dependent f. MRI signals. Journal of Cerebral Blood Flow and Metabolism : Official Journal of the International Society of Cerebral Blood Flow and Metabolism, 32(7), 1188– 1206. doi: 10. 1038/jcbfm. 2012. 23

Introduction Basic Principles Normal Anatomy BOLD Hardware/Software Paradigms Unusual Patterns Discussion At our institution we employ the following system -GE Discovery 750 W 3 Tesla MRI -Sensavue f. MRI by Invivo Corp. -Images processed in Dyna. Suite by Invivo Corp. Typical f. MRI Acquisition Parameters -TR = 2000 ms -Number of volumes 120 -Control 10 repetitions -Stimulus 10 repetitions Each volume takes 2 seconds to acquire and the first 4 volumes are discarded. Total scan time is 4 min 8 sec for each of our paradigms.

Introduction Basic Principles Normal Anatomy BOLD Hardware/Software Paradigms Unusual Patterns Discussion Acquisition Parameters, continued FSPGR BRAVO (IR-prep fast SPGR) image for anatomic localization Images are obtained using a 256 x 256 matrix with 1. 2 mm slice thickness. 1 slab is obtained with 16 locs per slab. TE = 3. 3, TR = 8. 5, NEX = 1, Flip Angle = 12 Diffusion Tensor Imaging (DTI) 4 mm Images are obtained using a 129 (freq. ) x 128 (phase) matrix with 5 mm slice thickness with 1 mm spacing. 29 slices are obtained. TEmin = 98. 6, TR = 8000, NEX = 1, B = 1000

Introduction Basic Principles Normal Anatomy Paradigms Language Motor Unusual Patterns Discussion _ Broca’s Area I Classically located in the left frontal lobe, refers to the pars triangularis(red) and the pars opercularis(blue) of the inferior frontal gyrus Wernicke’s Area Classically located in the posterior portion of the left superior temporal gyrus Arcuate Fasciculus A segment of the superior longitudinal fasciculus that connects Wernicke’s and Broca’s Image source: https: //en. wikipedia. org/wiki/Broca%27 s_area#mediaviewer/File: Broca%27 s_area_-_lateral_view. png. Licensed under CC-BY-SA-2. 1 -jp

Introduction Basic Principles Normal Anatomy Paradigms Language Motor _ Broca’s Area I Classically located in the left frontal lobe, refers to the pars triangularis and the pars opercularis of the inferior frontal gyrus Wernicke’s Area Classically located in the posterior portion of the left superior temporal gyrus Arcuate Fasciculus A segment of the superior longitudinal fasciculus that connects Wernicke’s and Broca’s Unusual Patterns Discussion

Introduction Basic Principles Normal Anatomy Paradigms Language Motor Unusual Patterns Discussion _ Broca’s Area I Classically located in the left frontal lobe, refers to the pars triangularis(red) and the pars opercularis(blue) of the inferior frontal gyrus Wernicke’s Area Classically located in the posterior portion of the left superior temporal gyrus Arcuate Fasciculus A segment of the superior longitudinal fasciculus that connects Wernicke’s and Broca’s Modified from: https: //en. wikipedia. org/wiki/Broca%27 s_area#mediaviewer/File: Broca%27 s_area_-_lateral_view. png. Licensed under CC-BY-SA-2. 1 -jp

Introduction Basic Principles Normal Anatomy Paradigms Language Motor Unusual Patterns Discussion _ Broca’s Area I Classically located in the left frontal lobe, refers to the pars triangularis and the pars opercularis of the inferior frontal gyrus Wernicke’s Area Classically located in the posterior portion of the left superior temporal gyrus Arcuate Fasciculus A segment of the superior longitudinal fasciculus that connects Wernicke’s and Broca’s Receptive language center In this healthy volunteer did not localize to the classic location

Introduction Basic Principles Normal Anatomy Paradigms Language Motor Unusual Patterns Discussion _ Broca’s Area I Classically located in the left frontal lobe, refers to the pars triangularis(red) and the pars opercularis(blue) of the inferior frontal gyrus Wernicke’s Area Classically located in the posterior portion of the left superior temporal gyrus Arcuate Fasciculus A segment of the superior longitudinal fasciculus that connects Wernicke’s and Broca’s Modified from: https: //en. wikipedia. org/wiki/Broca%27 s_area#mediaviewer/File: Broca%27 s_area_-_lateral_view. png. Licensed under CC-BY-SA-2. 1 -jp

Introduction Basic Principles Normal Anatomy Paradigms Language Motor _ Broca’s Area I Classically located in the left frontal lobe, refers to the pars triangularis(red) and the pars opercularis(blue) of the inferior frontal gyrus Wernicke’s Area Classically located in the posterior portion of the left superior temporal gyrus Arcuate Fasciculus A segment of the superior longitudinal fasciculus that connects Wernicke’s and Broca’s Unusual Patterns Discussion

Introduction Basic Principles Normal Anatomy Paradigms Language Motor Unusual Patterns Discussion _ Primary Motor Cortex I Classically located in the dorsal aspect of the precentral gyrus of the frontal lobe. Primary coordinator of movement. Premotor Cortex Located just anterior to the primary motor cortex. Function uncertain, may play a role in planning. Supplementary Motor Area Located at the medial surface of the hemispheres just anterior to the primary motor cortex. Contributes to control of movement. Image source: http: //www. neuroscientificallychallenged. com/glossary/primary-motor-cortex

Introduction Basic Principles Normal Anatomy Paradigms Language Motor _ Primary Motor Cortex I Classically located in the dorsal aspect of the precentral gyrus of the frontal lobe. Primary coordinator of movement. Premotor Cortex Located just anterior to the primary motor cortex. Function uncertain, may play a role in planning. Supplementary Motor Area Located at the medial surface of the hemispheres just anterior to the primary motor cortex. Contributes to control of movement. Unusual Patterns Discussion

Introduction Basic Principles Normal Anatomy Paradigms Language Motor Unusual Patterns Discussion _ Primary Motor Cortex I Classically located in the dorsal aspect of the precentral gyrus of the frontal lobe. Primary coordinator of movement. Premotor Cortex Located just anterior to the primary motor cortex. Function uncertain, may play a role in planning. Supplementary Motor Area Located at the medial surface of the hemispheres just anterior to the primary motor cortex. Contributes to control of movement. Image modified from: http: //www. neuroscientificallychallenged. com/glossary/premotor-cortex/

Introduction Basic Principles Normal Anatomy Paradigms Language Motor Unusual Patterns Discussion _ Primary Motor Cortex I Classically located in the dorsal aspect of the precentral gyrus of the frontal lobe. Primary coordinator of movement. Premotor Cortex Located just anterior to the primary motor cortex. Function uncertain, may play a role in planning. Supplementary Motor Area Located at the medial surface of the hemispheres just anterior to the primary motor cortex. Contributes to control of movement. Approximate location of premotor cortex highlighted in red

Introduction Basic Principles Normal Anatomy Paradigms Language Motor Unusual Patterns Discussion _ Primary Motor Cortex I Classically located in the dorsal aspect of the precentral gyrus of the frontal lobe. Primary coordinator of movement. Supplementary motor area Primary motor cortex Premotor Cortex Located just anterior to the primary motor cortex. Function uncertain, may play a role in planning. Supplementary Motor Area Located at the medial surface of the hemispheres just anterior to the primary motor cortex. Contributes to control of movement. Image modified from: http: //www. arts. uwaterloo. ca/~bfleming/psych 261/lec 16 no 9. htm

Introduction Basic Principles Normal Anatomy Paradigms Language Motor _ Primary Motor Cortex I Classically located in the dorsal aspect of the precentral gyrus of the frontal lobe. Primary coordinator of movement. Premotor Cortex Located just anterior to the primary motor cortex. Function uncertain, may play a role in planning. Supplementary Motor Area Located at the medial surface of the hemispheres just anterior to the primary motor cortex. Contributes to control of movement. Unusual Patterns Discussion

Introduction Basic Principles Normal Anatomy Paradigms Unusual Patterns Discussion Overview Motor Language Other _ f. MRI Paradigms -A paradigm is an activity or stimulus designed to elicit a specific cortical response -Typically target visual, speech, motor, or memory areas -Successful localization of eloquent areas requires that the patient is able to comply with the paradigm -Block design: the most common type during which a specific stimulus is repeated over a stimulus-rest cycle -Event related design: single events designed to elicit a cortical response ASFNR Paradigms. Available at http: //www. asfnr. org/paradigms. html. Date accessed 03/10/2015.

Introduction Basic Principles Normal Anatomy Paradigms Unusual Patterns Discussion Overview Motor Language Other Activation _ _time Block design paradigm schematic showing 20 second rest/stimulus cycles Stimulus Rest 20 40 60 80 100 120 Time (sec) Actual recorded data showing a block design paradigm with the smoothed predicted value and the actual recorded values (here labeled R 3)

Introduction Basic Principles Normal Anatomy Paradigms Unusual Patterns Discussion Overview Motor Language Other Motor Paradigms _ -Designed to stimulate motor, pre-motor and supplementary motor areas -Typically block design paradigms -Examples include: o o Bilateral complex finger tapping Unilateral sequential finger tapping Tongue movement/lip puckering Toe motion -Requires adequate patient dexterity and ability to cooperate (if lacking dexterity hand-grip can be attempted) -Physician or technologist can actively assist patient with finger motion or toe motion if patient unable to follow command in a timely fashion ASFNR Paradigms. Available at http: //www. asfnr. org/paradigms. html. Date accessed 03/10/2015.

Introduction Basic Principles Normal Anatomy Paradigms Unusual Patterns Discussion Overview Motor Language Other Motor Paradigms – Our methods _ -At our institution we begin every f. MRI with tongue movement to help separate this from the later language paradigms (if necessary) The patient is asked to move their tongue side to side in 20 second rest-stimulus intervals -Bilateral toe-motion The patient is asked to move their toes randomly in 20 second rest-stimulus intervals. It is important patients not move feet excessively to avoid head motion artifact -Bilateral finger tapping Rapid finger tapping over 20 second rest-stimulus cycles

Introduction Basic Principles Normal Anatomy Paradigms Unusual Patterns Discussion Overview Motor Language Other I Tongue movement (pink) with motor cortex activity Toe motion (yellow) with motor activity

Introduction Basic Principles Normal Anatomy Paradigms Unusual Patterns Discussion Overview Motor Language Other L _anguage Paradigms -Designed to lateralize and stimulate Broca’s area, Wernicke’s area and the arcuate fasciculus -Examples include: o o o Sentence completion Verb generation (i. e. Ball “throw”) Passive listening (Wernicke’s area) Word naming (i. e. think of words that start with “A”) Object naming -Typically a block design paradigm

Introduction Basic Principles Normal Anatomy Paradigms Unusual Patterns Discussion Overview Motor Language Other _ Language Paradigms -Requires the patient to be literate and able to recognize and read the letters on the display -It is essential that patients do not actually move their tongue/lips during the procedure to avoid confounding activation Gracco, V. L. , Tremblay, P. , & Pike, B. (2005). Imaging speech production using f. MRI. Neuroimage, 26(1), 294– 301. doi: 10. 1016/j. neuroimage. 2005. 01. 033

Introduction Basic Principles Normal Anatomy Paradigms Unusual Patterns Discussion Overview Motor Language Other Language Paradigms – Our methods _ We employ the following standardized paradigms: -Verb generation The patient is shown a noun and asked to think of a verb such as “car” “drive” -Word generation The patient is asked to think of as many words as they can that start with a specific letter -Sentence completion The patient is given a sentence and asked to complete it -Object naming The patient is shown various objects (e. g. house, chair) and asked to recognize and name them

Introduction Basic Principles Normal Anatomy Paradigms Unusual Patterns Discussion Overview Motor Language Other I Word generation (orange) showing activity in Broca’s area in a left language dominant subject. Sentence completion paradigm showing activity in Wernicke’s area (red)

Introduction Basic Principles Normal Anatomy Paradigms Unusual Patterns Discussion Overview Motor Language Other _ f. MRI Paradigms – Other uses -Script-driven f. MRI has been used to investigate patterns of activation in PTSD -Alzheimer’s disease characterization via parahippocampal and hippocampal activation in memory tasks -Assessment of addiction with focus on prefrontal cortex -Evaluation of traumatic brain injury (TBI), able to show subtle changes in mild TBI in the absence of structural defects on static images. Dependent on site of injury Hughes, K. C. , & Shin, L. M. (2011). Functional neuroimaging studies of post-traumatic stress disorder. Expert Review of Neurotherapeutics, 11(2), 275– 285. doi: 10. 1586/ern. 10. 198 Sperling, R. (2011). The potential of functional MRI as a biomarker in early Alzheimer's disease. Neurobiology of Aging, 32, S 37–S 43. doi: 10. 1016/j. neurobiolaging. 2011. 09. 009 Luijten, M. , Machielsen, M. , Veltman, D. , Hester, R. , de Haan, L. , & Franken, I. (2014). Systematic review of ERP and f. MRI studies investigating inhibitory control and error processing in people. Journal of Psychiatry & Neuroscience, 39(3), 149– 169. doi: 10. 1503/jpn. 130052 Mc. Donald, B. C. , Saykin, A. J. , & Mc. Allister, T. W. (2012). Functional MRI of mild traumatic brain injury (m. TBI): progress and perspectives from the first decade of studies. Brain Imaging and Behavior, 6(2), 193– 207. doi: 10. 1007/s 11682 -012 -9173 -4

Introduction Basic Principles Normal Anatomy Paradigms Unusual Patterns Discussion _Hx: 55 F Spanish-speaking with glioblastoma multiforme(red arrow) presents for preoperative evaluation prior to resection. Before starting the test it was discovered she was illiterate and thus unable to follow written directions. Presumed Wernicke’s area (blue arrow), slightly more anterior than expected, potentially distorted by tumor Challenge: Illiterate Spanish-speaking patient, unable to perform language tasks without assistance. Solution: Spanish-speaking interpreter read letters on screen and patient was able to complete word generation paradigm, resulting in faint activation of Wernicke’s area.

Introduction Basic Principles Normal Anatomy Paradigms Unusual Patterns Discussion _Hx: 51 M Cantonese-speaking with complex partial seizures due to a left frontal lobe mass(red arrow). Pre-operative evaluation given proximity to eloquent areas. Broca’s area localizes to inferior frontal lobule(blue arrow) Expected Wernicke’s area(yellow arrow) Object naming paradigm Challenge: Cantonese-speaking patient with limited English fluency. Solution: To complete object naming paradigm, the patient was shown images and asked to silently name them in his native language, resulting in activation of the language centers.

Introduction Basic Principles Normal Anatomy Paradigms Unusual Patterns Discussion _Hx: 37 F right-handed with intractable epilepsy s/p trauma found to have multiple prominent areas of left-sided activation during language tasks Two distinct Broca’s areas(blue arrows) Single Wernicke’s area(yellow arrow) Arcuate fasciculus Sentence completion paradigm Challenge: Multiple areas of activation on language related paradigm. Solution: Correlate with the patient’s medical record or patient query. We discovered the patient was trilingual, likely explaining the multiple Broca’s areas.

Introduction Basic Principles Normal Anatomy Paradigms Unusual Patterns Discussion _Hx: 71 M right-handed with intractable epilepsy for preoperative evaluation prior to right -sided amygdalohippocampectomy. Word generation paradigm Bilateral Broca’s areas (blue arrows) Challenge: Bilateral language centers with no clear dominant side by f. MRI. Evaluation for Wernickes was inconclusive. Solution: Confirmatory testing. Post-f. MRI preoperative Wada test confirmed bilateral codominant language centers, noting dysphasia with right-sided injection and non-fluency with left-sided injection. This patient suffered a postoperative right MCA territory infarct with hemorrhage resulting in severe aphasia consistent with Wada testing.

Introduction Basic Principles Normal Anatomy Paradigms Unusual Patterns Discussion Hx: _( 60 M left-handed semi-literate with right frontal glioblastoma multiforme(red arrow), preoperative evaluation. Faint Broca’s activation (blue arrow) DWI showing infarct Word generation paradigm Challenge: Semi-literate patient, able to recognize letters. Solution: The patient was asked to think of any words that came to mind when he saw a letter on the screen, which resulted in faint Broca’s area activation. This was confirmed by language deficits after postoperative infarct in this region.

Introduction Basic Principles Normal Anatomy Paradigms Unusual Patterns Discussion _Hx: 70 F right-handed with glioblastoma multiforme(red arrow) being evaluated prior to brain biopsy. f. MRI demonstrated robust tongue activation. Intense motor activity(blue arrow) Tongue movement paradigm Wernicke’s with arcuate fasciculus(yellow arrow) Word generation paradigm Challenge: Separate tongue activity from language activity in Wernicke’s area. Solution: Perform tongue movement paradigm first in order to localize tongue activity. This will help to differentiate activation of Wernicke’s area in case of inadvertent tongue movement during language generation paradigm.

Introduction Basic Principles Normal Anatomy Paradigms Unusual Patterns Discussion _Hx: 48 M with limited reading capacity, with large left-sided mass(red arrow) for evaluation prior to resection. f. MRI results influenced a change to conservative management. Intense motor activity(blue arrow)adjacent to mass Right and left corticospinal tracts(yellow arrows) Partially visualized left corticospinal tract(yellow arrow) Finger tapping paradigm Challenge 1: Left corticospinal tract not well visualized adjacent to mass. Solutions: Review literature and consult experienced colleagues. The mass caused medialization of the left corticospinal tract and adjacent edema likely altered diffusivity rendering the fibers unapparent at the level of the motor strip.

Introduction Basic Principles Normal Anatomy Paradigms Unusual Patterns Discussion large left-sided mass for evaluation prior to _Hx: 48 M with limited reading capacity, with to conservative management. resection. f. MRI results influenced a change Broca’s area(blue arrow) visualized using modified word generation paradigm Challenge 2: Patient had limited ability to recognize letters. Solution: Patient instructed to think of any word that came to mind regardless of letter presented in word generation paradigm.

Introduction Basic Principles Normal Anatomy Paradigms Unusual Patterns Discussion _Hx: 29 M, right handed, with intractable epilepsy and left mesial temporal sclerosis, evaluation prior to amygdalohippocampectomy. This patient was able to undergo leftsided resection without postoperative language deficits. Broca’s area(blue arrow) on the Wernicke’s area(yellow arrow) right & arcuate fasciculus on right Word generation paradigm Verb generation paradigm Challenge: Language center not in the expected location. Solution: Recognize normal anatomic variation. Approximately 5% of right handed patients can have right hemisphere language dominance. This was confirmed with a WADA test.

Introduction Basic Principles Normal Anatomy Paradigms Unusual Patterns Discussion -f. MRI is becoming a standard tool in the preoperative evaluation of neurosurgical patients and can affect operative planning -f. MRI must be performed and interpreted in the context of the patient with attention to pre-existing deficits (i. e. level of literacy, prior insult, ability to follow directions) -Unexpected patterns of activation can be challenging in a de novo f. MRI program.

Introduction Basic Principles Normal Anatomy Paradigms Unusual Patterns Discussion -Challenging cases we have encountered include bilateral, multiple or unexpected locations of language centers, and distortion of the usual anatomy. Additional history from the patient may be helpful. -Varying levels of literacy or English language proficiency that render implementation of paradigms challenging. The operator must improvise to complete the paradigm. Performing object naming tasks or involvement of interpreters can assist in paradigm completion. -Different paradigms are used to activate specific language centers(Brocas, Wernickes) however intensity and center of activation can be variable. Operator must utilize any available paradigms to obtain results useful for patient care. -When in doubt, refer to the literature and consult with experienced colleagues.