2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 1 Development of Complex Curricula for Molecular Bionics and Infobionics Programs within a consortial* framework** Consortium leader PETER PAZMANY CATHOLIC UNIVERSITY Consortium members SEMMELWEIS UNIVERSITY, DIALOG CAMPUS PUBLISHER The Project has been realised with the support of the European Union and has been co-financed by the European Social Fund *** **Molekuláris bionika és Infobionika Szakok tananyagának komplex fejlesztése konzorciumi keretben ***A projekt az Európai Unió támogatásával, az Európai Szociális Alap társfinanszírozásával valósul meg. PETER PAZMANY CATHOLIC UNIVERSITY SEMMELWEIS UNIVERSITY sote_logo.jpg dk_fejlec.gif INFOBLOKK 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 2 Biomedical Imaging fMRI–ClinicalApplications (Orvosbiológiai képalkotás) (fMRI–Klinikai alkalmazások) Lajos R. Kozák www.itk.ppke.hu Peter Pazmany Catholic University Faculty of Information Technology BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 3 Outline • General introduction to clinical fMRI• Goals, approaches, patient groups, paradigm selection • Introduction to clinical fMRIparadigms used in the MR Research Center (MRKK) at Semmelweis University, with example cases• Picture naming, synonym task, speech comprehension, auditory decision, memory encoding, home-town walking, sensory-motor task, retinotopicmapping • Specific issues in clinical fMRI• Single subject analysis, subject specific differences, pathology specific differences, lack of standardization • Validation specific issues• Effect of paradigm length, effect of smoothing, effect of thresholding, threshold-independent lateralization indices • Educationalcases• Cortical reorganization, post-surgical follow-up • Future applications• Connectivity mapping, pharmaceutical fMRI, BOLD and ASL mapping • Summary www.itk.ppke.hu BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 4 Introduction The main useof fMRIintheclinicalpracticeis theidentificantionof theso-calledeloquentareas, i.e. areasthatarenecessaryforpreservingqualityof life. • Sensory-motorcortex • Language-relatedareas• Broca • Wernicke • Visual cortex, etc. Thisgoalis, ingeneral, achievedbyusingtheprinciplesof brainmapping. www.itk.ppke.hu BiomedicalImaging: fMRI–ClinicalApplications The goal of pre-surgical fMRI To help guiding the scalpel of the neurosurgeon during neurosurgery, or the focus of radiation beams during radioablative therapy, while keeping as much function as possible. Clinical fMRI helps in decision making and treatment plannig to find the right trade-off between the maximal invasiveness of the intervention and the minimal loss of function. 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 5 www.itk.ppke.hu BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 6 Patients The main candidatesof pre-surgicalfMRIare: • Patientswithbraintumors • Patientswitharterio-venousmalformations • Patientswithdrug-resistantepilepsies • Patientswithmalformationsof corticaldevelopment • Patientswithdrug-resistantpainsyndromes www.itk.ppke.hu BiomedicalImaging: fMRI–ClinicalApplications ClinicalfMRIis notalwaysa stand-alonemethod Itis oftenusedinconjunctionwithotherfunctionalmappingapproaches, likeEEG/MEG and PET, dependingontheclinicalquestion. 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 7 Compared to EEG/MEG Advantage: • Precise spatial localization Disadvantage: • Worse temporal resolution • Much less flexible, there’s no bedside MRI (at the moment) Compared to PET Advantage: • Non-invasive, no ionizing radiation • More flexible paradigms can be used Disadvantage: • Deals with oxygenation only www.itk.ppke.hu BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 8 www.itk.ppke.hu Paradigm selection depends on the clinical question • In brain tumor patients the location of the lesion defines the focus and paradigm of mapping• Tumors near the central sulcus: sensory-motor cortex mapping • Tumors in the frontal of temporal regions: language mapping • Tumors in the occipital cortex: visual mapping • In drug-resistant epilepsy patients the clinical picture defines the paradigm and the approach• In case of a clearly defined epileptic focus the same is true as in brain tumors • In generalized epilepsies the identification of hemispherial language dominance is crucial • In epilepsies related to cortical malformations of development the identification of possible functional re-organizations can be helpful for treatment planning BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 9 Paradigms used in clinical fMRI • Are usually block-design paradigms• They provide the highest power in the shortest time • Relatively easy to explain to the patients • Tasks can be flexibly timed within blocks • The goal is to maximize functional contrast in the areas of interest while minimizing functional contrast in other areas• Well designed “passive”blocks contain no task related to the mapped functions, but contain tasks activatingunmapped areas:• picture naming task contains pictures in the “active” condition and the phase scrambled version of the same images during the “passive” conditions to minimize functional contrast in low level visual areas by providing the same luminance and spatial frequency components for both conditions • passive comprehension contains recorded speechin the “active” condition and the same recording reversed during the “passive” conditions to minimize functional contrast in low level auditory areas by providing the same frequency content for both conditions www.itk.ppke.hu BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 10 … … ACTIVE: Name it! Living/Object? CONTROL: Relax! Directionof arrows? >>> ClinicalMapping v6.6 © LR Kozák 2007-2010 , MRKK www.itk.ppke.hu ParadigmsusedintheMR Research Center (MRKK), Semmelweis University -Picture naming Duringtheactivepart of thetaskthepatienthas tocovertlynametheobjectpresentedontheimage and has tomakea living/objectdecision Duringthepassivepart thepatientis instructedtorelaxwithoutimagininganythingintothecloudyimage, and pressa buttonindicatingthedirectionof redarrows. Stimuli are presented in every 3s within 24s blocks. BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 11 The picture naming task activates the higher order visual areas (V), the Broca area (B) and the left premotor region (P, because of the required motor response) Patient examination @ MRKK in 2010, LR Kozák, MD, PhD V B P www.itk.ppke.hu ParadigmsusedintheMR Research Center (MRKK), Semmelweis University -Picture naming Patient1 Right temporal lobe epilepsy BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 12 ACTIVE: Synonym? CONTROL: Similar? door camel ZMVHB ZWVHB … … ClinicalMapping v6.6 © LR Kozák 2007-2010 , MRKK www.itk.ppke.hu ParadigmsusedintheMR Research Center (MRKK), Semmelweis University -Synonymtask Duringtheactivepart of thetaskthepatienthas toindicatebybuttonpresseswhetherthewordspresentedaresynonymsornot. Duringthepassivepart thepatienthas todecidewhetherthetwoconsonantstringsaresimilary, butis instructednottoreadtheletters. Stimuli are presented in every 3s within 24s blocks. BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 13 The synonym task activates the Broca area (B) and the left dorsolateral prefrontal cortex (D) and the left premotor region (P) Patient examination @ MRKK in 2010, LR Kozák, MD, PhD B D&P www.itk.ppke.hu ParadigmsusedintheMR Research Center (MRKK), Semmelweis University -Synonymtask Patient1 Right temporal lobe epilepsy BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 14 ACTIVE: Recorded speech CONTROL: Reversedspeech … … ClinicalMapping v6.6 © LR Kozák 2007-2010 , MRKK www.itk.ppke.hu ParadigmsusedintheMR Research Center (MRKK), Semmelweis University -Speechcomprehension Duringtheactivepart of thetaskthepatientis instructedtolistentoa pre-recordedspeech abouta neutraltopic(panda bears). Duringthepassivepart thepatientlistenstothesamerecordingmade incomprehensiblebyreversingit. Stimuli are presented in 24s blocks. Afterthescanningsession thepatientis askedsomequestionsaboutthespeechasa checkforattention. BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 15 The speech comprehension task activates Wernicke’s area (W) and the higher order auditory cortices (A). Patient examination @ MRKK in 2010, LR Kozák, MD, PhD W A www.itk.ppke.hu ParadigmsusedintheMR Research Center (MRKK), Semmelweis University -Speechcomprehension Patient1 Right temporal lobe epilepsy BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 16 ACTIVE: Words vs. non-words CONTROL: Beepswithdifferentfrequencies … … ClinicalMapping v6.6 © LR Kozák 2007-2010 , MRKK www.itk.ppke.hu ParadigmsusedintheMR Research Center (MRKK), Semmelweis University -Word-pseudowordtask Duringtheactivepart of thetaskthepatientis instructedtomakeword-pseudoworddecisiononthepresentedHungarianwords/pseudowords. Duringthepassivepart thepatientis instructedtomakedecisiononthepitchof beepspresented. Stimuli are presented in every 3s within 24s blocks. 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 17 The word-pseudoword task activates Wernicke’s area (W), the higher order auditory cortices (A), and Broca’s area (B). Patient examination @ MRKK in 2010, LR Kozák, MD, PhD W A B www.itk.ppke.hu ParadigmsusedintheMR Research Center (MRKK), Semmelweis University -Word-pseudowordtask Patient1 Right temporal lobe epilepsy BiomedicalImaging: fMRI–ClinicalApplications ACTIVE: Try to remember CONTROL: Relax … … 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 18 ClinicalMapping v6.6 © LR Kozák 2007-2010 , MRKK www.itk.ppke.hu ParadigmsusedintheMR Research Center (MRKK), Semmelweis University -Memory encoding Duringtheactivepart of thetaskthepatientis instructedtolook at the images and try to memorize them. The whole set is presented twice. Duringthepassivepart thepatientis instructedto relax. The image pool contains 60 images. Stimuli are presented in every 3s within 30s blocks. After the scanning session 32 images is shown to the patient who has to indicate which of them were presented previously. Ávilaet al. Am J Neurorad, 2006 BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 19 The task activates various areas including the visual cortex, areas involved in visual attention, even the Broca area. The cross shows a left lateralized activation focus in the temporal white matter. Patient examination @ MRKK in 2010, LR Kozák, MD, PhD www.itk.ppke.hu ParadigmsusedintheMR Research Center (MRKK), Semmelweis University –Memory encoding Patient1 Right temporal lobe epilepsy BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 20 ACTIVE: Imagine the route CONTROL: Count FromHOME ToPOST OFFICE Up from 21by 2 … … ClinicalMapping v6.6 © LR Kozák 2007-2010 , MRKK www.itk.ppke.hu ParadigmsusedintheMR Research Center (MRKK), Semmelweis University –Hometown walking Duringtheactivepart of thetaskthepatientis instructedtoimaginewalkingalonga familiarroute, and tovisualizethesurroundings. Duringthepassivepart thepatientis instructedtocountaccordingtothegiveninstruction. Stimulationis donein 30s blocks. Ávilaet al. Am J Neurorad, 2006 BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 21 The task activates various areas. The cross shows an activation focus in the left mesial temporal lobe that is more extensive than that of the right mesial temporal lobe, suggestive of left lateralization of memory retrieval. Patient examination @ MRKK in 2010, LR Kozák, MD, PhD www.itk.ppke.hu ParadigmsusedintheMR Research Center (MRKK), Semmelweis University –Hometown walking Patient1 Right temporal lobe epilepsy BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 22 ACTIVE: Move the indicated limb CONTROL: Rest <<< LEFT REST … … ClinicalMapping v6.6 © LR Kozák 2007-2010 , MRKK www.itk.ppke.hu ParadigmsusedintheMR Research Center (MRKK), Semmelweis University -Sensory-motor mapping Duringtheactivepart of thetaskthepatientis instructed to move the indicated limb. Hand areas are mapped by thumb opposition tasks; feet areas are mapped by a toe movement tasks; face areas are mapped by tongue movement task. Duringthepassivepart thepatientis instructed to rest passively. 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 23 The three tasks map the sensory-motor region along the central sulcus (marked with red line). Patient examination @ MRKK in 2009, LR Kozák, MD, PhD toe finger tongue www.itk.ppke.hu ParadigmsusedintheMR Research Center (MRKK), Semmelweis University –Sensory-motor mapping Patient2 Motor cortex mappingin drug resistant pain syndrome BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 24 The three tasks mapped the sensory-motor region along the central sulcus. The mapping opened the possibility for minimally invasive electrode implantation in a stereotactic setting, that resulted in 60% decrease in perceived pain intensity. Patient examination @ MRKK in 2009, LR Kozák, MD, PhD Intraoperative images courtesy of I Valálik MD Department of Neurosurgery, Szt János Kórház, Budapest, Hungary DSC01691 DSC01698 www.itk.ppke.hu ParadigmsusedintheMR Research Center (MRKK), Semmelweis University –Sensory-motor mapping Patient2 Motor cortex mappingin drug resistant pain syndrome BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 25 Polar angle mapping: Eccentricity mapping: <<< LEFT RET ClinicalMapping v6.6 © LR Kozák 2007-2010 , MRKK www.itk.ppke.hu ParadigmsusedintheMR Research Center (MRKK), Semmelweis University –Retinotopicmapping During retinotopicmapping a polar coordinate system representation of the visual field is fitted to the retinotopicvisual areas. The mapping consists of two steps: polar angle mapping by a rotating wedge stimulus, and eccentricity mapping by an extending ring stimulus. Both stimuli have a superimposed counterphasing(8Hz) checkerboard pattern. BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 26 The dysgenesis (marked in green on the top right image) does not interfere with visual processing in the retinotopic visual areas (bottom images). Patient examination @ MRKK in 2008, LR Kozák, MD, PhD www.itk.ppke.hu ParadigmsusedintheMR Research Center (MRKK), Semmelweis University –Retinotopicmapping Patient3 Retinotopicmapping in a case of occipital cortical dysgenesis BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 27 www.itk.ppke.hu Paradigmselectiondependsonthepatient, aswell Patientscan’talwaysperformthetasksasintended • The taskis toocomplicatedfortheage, IQ, education, etc• The solutionis simplification:• Leavingout attentionaltask • Leavingout taskonpassivecondition • Usingwordstogeneratesentences • Usingletterstogeneratewords • The patientcan’tseeorhear• Changestimulusmodality • The patientcan’tmove• Asktoimaginemovement • Dopassivemovement, even in sedationSouweidaneet al., 1999 PediatrNeurosurg; Liu et al., 2005 Br J Anaesth, Kozaket. al Symp. Neurorad, 2010 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 28 www.itk.ppke.hu SpecificissuesinclinicalfMRIanalysis WhileresearchorientedfMRIstudies(includingclinicalresearch, aswell) areusuallygroupstudieswithgroupslevelinferences, clinicalfMRIstudiesareusuallyanalyzedonthesinglesubjectlevel. • WhileresearchorientedfMRIanalysesdealswiththemultiplecomparisonproblembylimiting thenumberof falsepositives• Bonferronicorrection • Falsediscoveryrate • Familywiseerror • Ina singlesubjectanalysislimiting falsenegativevoxelsmightequallybe important• Usinga more liberalstatisticalthresholdwithclustersizethresholding• Butthisraisesfurtherquestions BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 29 www.itk.ppke.hu SpecificissuesinclinicalfMRIanalysiscont’d Everyone “works” in a different way Although the shape of the hemodynamic response is roughly similar among functional areas,Boynton et al., J Neurosci, 1996; Josephs et al., HBM, 1997, Zarahnet al., NeuroImage, 1997 response dynamics are different across brain regionsSchacteret al., NeuroImage, 1997 and individuals.Aguirre et al., NeuroImage, 1998 BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 30 The pattern of activations depends heavily on the state of the patient (alertness, attention, anxiety, medications taken, etc.) In the experiment of McGonigleet al. the same subject performed the same task 33 times in a two-months period. The activation maps differed substantially betweens sessions. Proper pre-processing can limit the inter-session variability. Smith et al. HBM, 2005 www.itk.ppke.hu Specific issues in clinical fMRI analysis cont’d McGonigle et al., NeuroImage, 2000 BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 31 www.itk.ppke.hu SpecificissuesinclinicalfMRIanalysiscont’d Patients’ state and BOLD signal: • Everything vasodilatator: signal • hyperventillation(e.g. stress related) • administration of insulin in diabetics • Anaemia • Everything vasoconstrictor: signal • hypercapnia • theophyllin/ caffeine • high hematocrit • There are cycle-specific effects in femalessignal BiomedicalImaging: fMRI–ClinicalApplications The activation maps differ across individuals The localization of language areas are very variable across individuals Binder et al., J Neurosci 1997; Stippich et al., Neurosci Lett, 2003 Cognitive functions (thus brain responses) are age-dependentRotte et al., Age and Ageing, 2005 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 32 Probabilistic map of a picture naming task The more patients activate a given voxel in Talairach normalized space the brighter the color representation is.(Normative data from MRKK, LR Kozak et al., ESNR 2008) www.itk.ppke.hu Specific issues in clinical fMRI analysiscont’d BiomedicalImaging: fMRI–ClinicalApplications The activation maps differ betweenparadigms Language lateralization depends on the paradigms used Carpentier et al., Epilepsia, 2001; Baciu et al., Neuroradiol 2005 Language maps depend on the paradigms used Kozak et al., ESNR, 2008 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 33 Language activations in the Broca area depend on the paradigm used Picture naming;Synonym task;Intersection (Normative data from MRKK, LR Kozak et al., ESNR 2008) www.itk.ppke.hu Specific issues in clinical fMRI analysiscont’d BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 34 www.itk.ppke.hu SpecificissuesinclinicalfMRIanalysiscont’d The BOLD response depends on brain pathology e.g. in the vicinity of large gliomas, the BOLD amplitude decreases at least in about half of the cases Grummichet al., NeuroImage, 2006 Lesion-related changes might stem from: • Compression signal • Neovascularizationsignal • Metabolic changes signal • Therapy (drugs, surgery) signal • Cavernous angioma(susceptibility)signal • Epileptic activitysignal BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 35 The cavernous angioma in the temporo-parieto-occipital junction causes an extensive signal dropout in the BOLD-EPI images near the expected location of the Wernicke area. Patient examination @ MRKK in 2010, LR Kozák, MD, PhD www.itk.ppke.hu Specific issues in clinical fMRI analysis cont’d Patient 4 Cavernous angioma BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 36 The cavernous angioma in the temporo-parieto-occipital junction causes an extensive signal dropout in the BOLD-EPI images. The Wernicke area can’t be mapped in this patient despite the lack of apparent language deficit. The cross shows the Broca area which is not affected by the susceptibility artifact caused by cavernous angioma. Patient examination @ MRKK in 2010, LR Kozák, MD, PhD www.itk.ppke.hu Specific issues in clinical fMRI analysis cont’d Patient 4 Cavernous angioma BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 37 The big frontal tumor compresses the inferior frontal gyrus (IFG), the anatomical region where Broca area is expected. Patient examination @ MRKK in 2010, LR Kozák, MD, PhD s. centr s. centr s. centr s. precentr s. precentr s. precentr s. centr s. precentr IFG p. operc IFG p. operc IFG p. operc IFG p. operc insula insula IFG p. triang. IFG p. triang. www.itk.ppke.hu Specific issues in clinical fMRI analysis cont’d Patient 5 Frontal tumor BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 38 The big frontal tumor compresses the inferior frontal gyrus (IFG), the anatomical region where Broca area is expected. The activation at the Broca area is less extensive than in normal controls. Patient examination @ MRKK in 2010, LR Kozák, MD, PhD Pict. NSynonSpeechWrd/NW www.itk.ppke.hu Specific issues in clinical fMRI analysis cont’d Patient 5 Frontal tumor BiomedicalImaging: fMRI–ClinicalApplications Patient6 Polymicrogyriawith drug resistant epilepsy Epileptic activity can seriuoslyaffect fMRI In a case of cortical dysgenesisin a pediatric patient we encountered a condition of electric status epilepticusduring sleep (ESES) upon propofolanesthesia. As the amplitude of epileptic activity (700µV) was more than 10 times higher than the expected 5µV amplitude of the somatosensoryevoked potentials with propofolanesthesia (Liu et al. Br J Anaesth, 2005) ESES masked the effect of passive limb movement. Patient examination @ MRKK in 2008 Kozaket al., IdeggyogySz, 2009 Kozaket al., Symposium Neuroradiologicum, 2010 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 39 With Clonazepam Without Clonazepam www.itk.ppke.hu Specific issues in clinical fMRI analysis cont’d BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 40 www.itk.ppke.hu SpecificissuesinclinicalfMRIanalysiscont’d The lack of standardization There is still a lack of standardization regarding paradigms, processing steps, statistical methods • This is partly due to differences in equipment • Differences in clinical practice Currently the only solution is to create in-house normative databases • Evaluate the paradigms on healthy subjects prior to patients • Re-evaluate the paradigms based on patient studies • Re-evaluate the paradigms based on input from neurologists and neurosurgeons BiomedicalImaging: fMRI–ClinicalApplications Specific issues in in-house validation The activation maps depend on the number of blocks As the number of stimulation blocks increases the signal to noise ratio also increases. The statistical maps become more and more detailed as more voxelssurvive the multiple comparison correction at a given statistical significance level. Kozaket al., ESNR, 2008 Kozaket al., Symposium Neuroradiologicum, 2010 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 41 4 5 6 7 8 9 10 11 12 www.itk.ppke.hu BiomedicalImaging: fMRI–ClinicalApplications The activation maps depend on preprocessing parameters Spatial smoothing increases the signal to noise ratio. The statistical maps become more widespread with smoothing as more voxelssurvive the multiple comparison correction at a given statistical significance level.(However, spatialresolutionis decreasingwithincreasingsmoothingkernel) Kozaket al., ESNR, 2008 Kozaket al., Symposium Neuroradiologicum, 2010 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 42 0 mm 4 mm 8 mm 12 mm www.itk.ppke.hu Specific issues in in-house validation cont’d BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 43 FDR q<0.05 FDR q<0.01 FDR q<0.005 FDR q<0.001 Bonf p<0.05 Bonf p<0.01 Bonf p<0.005 Bonf p<0.001 www.itk.ppke.hu Specificissuesinin-house validation cont’d Statistical thresholdingdetermines the activation map. With stricter thresholds the number of false positives decrease, thus the extent of activations also decrease. Kozaket al., ESNR, 2008; Kozaket al., Symposium Neuroradiologicum, 2010 BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 44 www.itk.ppke.hu SpecificissuesinclinicalfMRIanalysiscont’d Lateralizationindex calculation Importantingeneralizedepilepsiestoasseslanguagelateralization. LI=(LeftActiveVoxels-RightActiveVoxels)/(LeftActiveVoxels+RightActiveVoxels) Asthestatisticalmapsheavilydependonthresholdinga novelthresholdindependentmethodforlanguagelateralizationestimationwassuggestedbySuarezet al. (EpilepsyBehav, 2009). Theyapproachis basedontheweighteddistributionof t-scoresfoundintheROIs. The thresholdindependLI calculationleadstothesameresultsasFDR q<0.05 thresholding. Tóth et al. & Kozaket al. Symp. Neurorad., 2010 BiomedicalImaging: fMRI–ClinicalApplications Patient6 Polymicrogyriawith drug resistant epilepsy Activation upon passive right hand movement(healthylimb)in propofolsedation shows up in the expected location in the contralateral(healthy)hemisphere, both preoperatively and postoperatively Patient examination @ MRKK in 2008 Kozaket al., IdeggyogySz, 2009 Kozaket al., Symp. Neuroradiologicum, 2010 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 45 www.itk.ppke.hu fMRI can prove functional reorganization BiomedicalImaging: fMRI–ClinicalApplications Patient6 Polymicrogyriawith drug resistant epilepsy Activation upon passive lefthand movement (affectedlimb) in propofolsedation shows up in theipsilateral(healthy)hemisphere, both preoperatively and postoperatively, suggestiveof functionalreorganizationtothehealthyhemisphere Patient examination @ MRKK in 2008 Kozaket al., IdeggyogySz, 2009 Kozaket al., Symp. Neuroradiologicum, 2010 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 46 www.itk.ppke.hu fMRI can prove functional reorganization Patient7 Precentraltumor Right finger tapping activations (shown in greenish blue) and left finger tapping activations (shown in yellow) in a case of precentraltumor. The left hand activations were present on the posterior edge of the lesion, so fMRIalone was not safe enough to delineate functionally active areas, therefore intraoperativeelectrocorticalstimulation was also applied for motor cortex mapping. Patient examination @ MRKK in 2009 LR Kozak, MD, PhD 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 47 www.itk.ppke.hu fMRI can be used pre-and postoperatively BiomedicalImaging: fMRI –ClinicalApplications Patient7 Precentraltumor Right finger tapping activations (shown in greenish blue) and left finger tapping activations (shown in yellow) in the half year follow-up examination of precentraltumor. Left hand activations shown posterior to the scar, seem to be normal. The fMRIfinding is supported by the fact that the patient had intact hand movement capabilities post-op. Patient examination @ MRKK in 2009 LR Kozak, MD, PhD 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 48 www.itk.ppke.hu fMRI can be used pre-and postoperatively BiomedicalImaging: fMRI –ClinicalApplications BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 49 Futureapplications Thereis widespreadresearchgoingonforextendingthepossibilitiesof clinicalapplicationsof fMRI, theseinvestigationsinclude, butarenotlimited to, thefollowing: • Functionalconnectivityanalysisincasesof epilepsy, dementias, etc. e.g. Bettuset al., JNNP, 2010 • CalibratedfMRI e.g. Mark et al., NeuroImage, 2010 • Crossvalidationof ASL perfusionimaging, BOLD fMRIand othermethods e.g. Diekhoffet al., HBM, 2010 • Estimationof drugeffectswithBOLD fMRI e.g. Luiet al., ArchGenPsy, 2010 • fMRI-basedcomplexbiomarkerresearch e.g. Paulsenet al., AJNR, 2004 www.itk.ppke.hu BiomedicalImaging: fMRI–ClinicalApplications 2011.10.04.. TÁMOP –4.1.2-08/2/A/KMR-2009-0006 50 Summary The introduced paradigms and instructive cases provide a comprehensive overview of the current clinical applications, but clinical fMRIis not limited to pre-surgical workup. Moreover, research related to clinical fMRIare not limited to methodological investigations, as clinically oriented research may use fMRIas a tool for assessing cognitive or other functional changes in patient groups compared to healthy individuals. Such research applications may lead to clinically important cut-off values, or complex fMRI-based biomarkers that can later be integral to routine diagnostic or prognostic procedures. www.itk.ppke.hu