SS 311b - Functional MRI
Independent component and seed-based analyses in localisation and lateralisation of Broca's and Wernicke's areas with resting state fMRI
Purpose: To compare SCA and ICA methods in mapping Broca's and Wernicke’s areas with resting-state fMRI (rsfMRI) data in healthy volunteers.
Methods and Materials: Each of 52 volunteers (25 left-handed) performed two task-based and one rsfMRI session in the Siemens Avanto 1.5T scanner. Each of the Broca’s and Wernicke’s areas was localised by three methods: using one of the speech tasks (SPM12 software); ICA analysis of rsfMRI data and the language network mask (GIFT v4.0); and SCA analysis of the rsfMRI data with a seed region in the other speech area revealed by the independent speech task (Conn v.14). Location and laterality index (LI) of the speech areas identified with the two rsfMRI methods were cross-validated with the task-based data.
Results: The average overlap of the speech areas localized by task-based and rsfMRI (either SCA or ICA) has not exceeded 25%. However, correlation between LI across SCA and task-based analyses was moderate for Broca’s area (r=0.67, p<0.001) and high in Wernicke's area (r=0.71, p<0.001). The ICA showed no LI correlation with the task-based data: r=0.26 (p=0.06) for Broca's area and r=0.08 (p=0.57) for Wernicke's.
Conclusion: Unlike task-based and SCA-based results, the speech areas localized by ICA analysis of rsfMRI were bilateral in all volunteers. Therefore, speech areas are better lateralized by SCA method of rsfMRI rather than ICA. It is still to be empirically tested whether the flexible adjustment of the statistical thresholds on the individual basis may improve the consistency of task-based and rsfMRI data.
Purpose: To perform functional connectivity analysis based on resting-state fMRI data for differential diagnostics between vegetative (VS) and minimal consciousness states (MCS).
Methods and Materials: In 14 patients with chronic disorders of consciousness (DOC) (CRS-rVS=4, CRS-rMCS-=5, CRS-rMCS+=5, mean age 26+-2.0) rsfMRI (3 T Siemens Verio) was performed followed by data preprocessing in SPM8 and connectome construction in CONN. As regions of interest (ROI) we accepted full standard brain atlas. We obtained connectome with functional connections showing correlation of BOLD signal in ROIs. Also, we analysed correlation coefficients in default mode network's (DMN) nodes. Statistical results were thresholded using a combination of a connection-level threshold (FDR-corrected p values on individual ROI-to-ROI connections) and seed-level thresholds (F test on multivariate connectivity strength for each seed ROI).
Results: MCS- group showed appearance of three connections with a positive correlation, and MCS + group showed significant increase of the amount of connections. Improving of the level of awareness is associated with increase of the correlation coefficient between medial prefrontal cortex (MPFC) only and other areas outside the DMN. Values of correlation coefficients between MPFC and left lateral parietal cortex were: -0.01 (VS), 0.05 (MCS-), 0.05 (MCS+); between MPFC and posterior cingulate cortex: -0.01 (VS), 0.1 (MCS-), 0.1 (MCS+); between MPFC and right lateral parietal cortex: -0.12 (VS), 0.1 (MCS-), 0.14 (MCS+).
Conclusion: The connectoms' number of connections in ROI multiplies with the level of consciousness increase. MPFC may be of most important value for further investigations because of its relations with other DMN zones.
Purpose: Pre-surgical functional MRI (fMRI) is rapidly becoming a standard method to detect the eloquent brain areas. Breathing and pulsation cause significant artefacts, which are decreasing the localisation sensitivity and specificity. We would like to demonstrate the effectivity of physiological artefact removal - based on physiological parameter recordings - on the sensitivity and selectivity of the fMRI mapping.
Methods and Materials: Siemens Magnetom Verio 3T MRI system was used to collect data. The physiological parameters (breathing, pulse and ECG) were recorded with the built-in devices. 20 patients - with brain tumour - data were evaluated with SPM12. RETROICOR/RVHR was applied to detect the relevant physiological artefacts. We compared the statistical maps before and after the physiological correction.
Results: The physiological correction improved the specificity: we found false-positive activation around the main brain vessels. The main relative activation value of the eloquent areas did not change significantly after the correction, albeit the standard deviations were lower. The correction improved the statistical maps in 11 patients. We detected minimal changes in 5 patients as a result of low pulse and slow breathing.
Conclusion: The physiological correction increased the specificity of statistical maps, so we are able to detect the eloquent regions more reliably, which allows more reliable pre-surgical planning.
Repeated application of realtime-fMRI neurofeedback in tobacco dependent patients after smoking cessation
Purpose: To evaluate the effect of repeated application of real-time(rt)-fmri neurofeedback in patients with tobacco dependency with respect to smoking cessation success.
Methods and Materials: Tobacco-dependent patients (n=43) were randomized into two groups receiving either real feedback (n=26) of an addiction associated brain region (ACC, insula, DLPFC) or sham feedback. The aim was to reduce neural activity in the region of interest during nicotine cue exposure in 3 sessions within a period of 4 weeks after professionally assisted smoking cessation. Preprocessing and statistical analysis of functional images were conducted in BrainVoyagerQX taking patients’ smoking status after 6 months into account (persistent tobacco abstinence/ significant reduction or complete relapse). Clinical data were assessed by established questionnaires (QSU, BDI, BIS, FTND).
Results: The fixed effect analysis of all patients with known smoking status after 6 months (real feedback, group) revealed brain activity reduction in addiction associated brain region in both abstinent/ less smoking patients (n=17) and completely relapsed patients (n=7) during neurofeedback sessions (3rd versus 1st run, p<0.05, Bonferroni corrected). However, in contrast to relapsed patients, abstinent/ less smoking patients had reduced brain activity in the ACC and the nucleus caudatus in several runs and enhanced brain activity in thalamic and parietal regions, especially in the 2rd and 3rd run of each session (between-group-comparison, p<0.05, Bonferroni corrected).
Conclusion: Neurofeedback based on rt-fmri may reveal and induce specific anti-addiction changes of brain state in tobacco dependent patients after smoking cessation and probably be used as a diagnostic and therapeutic tool.
Purpose: To identify executive function deficits assessed in functional magnetic resonance imaging (fMRI) studies of bipolar disorder (BD) through a voxel-based meta-analysis.
Methods and Materials: Web-based publication databases were searched to perform this meta-analysis of whole brain fMRI studies up to July 2016. A new improved voxel-based meta-analytic method, signed differential mapping (SDM), was developed to examine regions of activated or deactivated grey matter regions in the BD patients vs. healthy controls during a variety of executive functioning tasks.
Results: Eighteen datasets comprising 461 people with BD and 423 healthy controls met the inclusion criteria. Compared with age-matched healthy control participants, participants with BD showed reliable patterns of hyperactivation in bilateral gyrus rectus, extending to bilateral anterior cingulate cortex (ACC) and right middle temporal gyrus; hypoactivation in left cerebellum crus I, left precentral gyrus, right precuneus, right dorsolateral prefrontal gyrus (DLPFC). A sensitivity analysis as well as analyses of subgroups further confirmed these findings. Meta-regression analyses showed that studies that included individuals with older age of onset were significantly more likely to report hyperactivation area in the right ACC.
Conclusion: These findings suggest that BD pts show greater activation mainly in the prefronto-temporal gyrus during executive function tasks which could represent a major neural functional correlate of their illness as defined in DSM-V.
Stroke-related brain plasticity of the human mirror-neuron system during audio-motor transformation: fMRI and DTI study
Purpose: Audio-motor transformation (AMT) is a good model for perception to action information flow. Human mirror-neuron system (MNS) consisted of ventral premotor and inferior parietal cortices and was shown to participate in AMT. We propose brain activation, structural and functional connectivity analysis for the focal-stroke-related plasticity changes in AMT.
Methods and Materials: Three groups (G1, G2, G3) of right-handed subjects were studied. G1: 5 healthy subjects (3F, 45-63 y.o.), G2: 3 primary sensorimotor stroke patients (1F, 41-76 y.o.), G3: 2 left supramarginal gyrus stroke patients (1F, 44-58 y.o.). We used 1.5T SIGNA (GE, USA) for DTI and fMRI. Finger tapping was used for activation. EPI was used for BOLD imaging (TR/TE=3000/71ms, voxel 4x4x5mm) and DTI (b=1000, TR/TE=8000/100ms, voxel 0.9x0.9x4mm, encoded directions =25). Data processing was done using FSL (Oxford, GB) and BrainWavePA (GE, USA).
Results: Decrease of FA (G3=0.08, G1=0.4) and fiber density was shown for G3 at ischaemic focus location. Altered AMT-related brain activation was found in G3, in comparison to G1, G2. G1, G2 AMT activation pattern: primary sensorimotor cortex, supplementary motor area, bilateral temporal cortex, suparamarginal, inferior frontal cortices (MNS). G3 showed altered AMT-related activation: lesion nearby left supramarginal gyrus slight activation, bilateral temporal cortices. Thus, one of the mirror-neuron nodes' impairment led to the decrease of MNS AMT functioning, resulting in decreased auditory stimulus response and voice command execution.
Conclusion: fMRI and DTI could reveal stroke-related brain plasticity. Left supramarginal gyrus plays important role in MNS functioning during AMT.
Functional and structural changes in brain’s default mode network in early stages Parkinson’s disease patients according to voxel-based morphometry and resting-state fMRI comparison
Purpose: To assess pattern of functional and structural changes in neurodegeneration comparing to physical examination and healthy controls by resting-state fMRI (RsfMRI) and voxel-based MRI (VBM) is very prospective nowadays study, especially changes within default mode network (DMN).
Methods and Materials: We have exanimated, compared 3 independent right-handed groups - 41 patients with PD in early steps in pharmacological treatment and newly diagnosed, 20 healthy volunteers were investigated summary and underwent 1.5 T RsfMRI, T1-MPR MRI scanning for fMRI, VBM data acquisition.
Results: We had some results (p<0,005): in group of newly diagnosed PD patients we observed significant greater volume of right precuneus (part of DMN) comparing to patients in pharmacological treatment and healthy volunteers, in group of patients in treatment we marked volume loss in right precuneus comparing to newly diagnosed PD patients and healthy volunteers. Comparing fMRI data and volume changes in precuneus in group of newly diagnosed PD patients we observed direct correlation (than higher volume of precuneus that greater areas of spontaneous neuronal activity in precuneus), in group of patients in treatment we observed inversed correlation (than smaller precuneus volume that significant greater areas of spontaneous neuronal activity in precuneus).
Conclusion: Our findings may indicate about compensatory hypertrophia of precuneus as compensatory mechanisms of neuroplasticity phenomena in early stages of PD, at that time volume loss of precuneus in PD patients in treatment may reflect increasing atrophy in neurodegenerative process. Correlation between fMRI and VBM-analysis data may be demonstration of multivariants of compensatory mechanisms in neurodegeneration.
Purpose: Mirror therapy is an effective rehabilitation therapy after stroke. A mirror is placed between the patient's arms or legs so that the image of the non-affected limb gives the illusion of normal movement in the affected limb. Despite the overwhelming evidence of adult brain’s capacity to reorganize after stroke, it remains unclear whether and to what extent specific peri-lesional or full brain adaptation is responsible for motor-function recovery.
Methods and Materials: In a randomised study, starting with either conventional or Mirrorbox Therapy, we used a repetitive task specific approach (forearm, wrist and hand) 3x15 minutes/day) for 3 weeks at which time patients switched therapy group. Resting-state-fMRI was performed at baseline, crossover (3 weeks) and 6 weeks.
Results: Resting-state fMRI showed changes in inter-hemispheric synchronization of neuronal signalling and reorganization of the motor execution network after stroke rehabilitation. This was more pronounced after Mirror therapy compared to conventional rehabilitation. There was specific activation of the temporal gyrus, precuneus and the posterior cingulate cortex (PCC).
Conclusion: It appears that alterations in functional brain networks are interrelated to the structural plasticity of the underlying neuronal architecture. Activation of the temporal gyrus, precuneus and the PCC might be related to improved self-awareness and spatial attention.