Lecture10_GroupICA

Lecture10_GroupICA - fMRI Course Lecture 10 Group ICA Vince D Calhoun Ph.D Director Image Analysis& MR Research The MIND Institute Associate

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: fMRI Course Lecture 10: Group ICA Vince D. Calhoun, Ph.D. Director, Image Analysis & MR Research The MIND Institute Associate Professor, Electrical and Computer Engineering, Neurosciences, and Computer Science The University of New Mexico Introduction Introduction • Using ICA to analyze fMRI data of multiple subjects raises some questions: • How are components to be combined across subjects? • How should the final results be thresholded and/or presented? 2009 fMRI Course 2 Group ICA Group a Combine Single Subject ICA’s1,4 Unique Spatial Unique Temporal Group ICA Approaches Group b Temporal Concatenation3,7,5 Common Spatial Unique Temporal Voxels Correlate/Cluster Time c Spatial Concatenation6,5 Unique Spatial Common Temporal Voxels d Pre-Averaging5 e Common Spatial Common Temporal Tensor2,7 Common Spatial Common Temporal Subject Parameter Voxels ICA ICA Subject N } Time Subject 1 : Subject N Time : Sub 1 Subject 1 Subject 1 Subject N Subject (avg) ? Back reconstruction } GIFT Single subject maps Single subject components* MELODIC Brain Voyager 1) Calhoun VD, Adali T, McGinty V, Pekar JJ, Watson T, Pearlson GD. (2001): fMRI Activation In A Visual-Perception Task: Network Of Areas Detected Using The General Linear Model And Independent Component Analysis. NeuroImage 14(5):1080-1088. 2) Beckmann CF, Smith SM. (2005): Tensorial extensions of independent component analysis for multisubject FMRI analysis. NeuroImage 25(1):294-311. 3) Calhoun VD, Adali T, Pearlson GD, Pekar JJ. (2001): A Method for Making Group Inferences from Functional MRI Data Using Independent Component Analysis. Hum.Brain Map. 14(3):140-151. 4) Esposito F, Scarabino T, Hyvarinen A, Himberg J, Formisano E, Comani S, Tedeschi G, Goebel R, Seifritz E, Di SF. (2005): Independent component analysis of fMRI group studies by self-organizing clustering. Neuroimage. 25(1):193-205. 5) Schmithorst VJ, Holland SK. (2004): Comparison of three methods for generating group statistical inferences from independent component analysis of functional magnetic resonance imaging data. J.Magn Reson.Imaging 19(3):365-368. 6) Svensen M, Kruggel F, Benali H. (2002): ICA of fMRI Group Study Data. NeuroImage 16:551-563. 7) Guo Y, Giuseppe P. (In Press): A unified framework for group independent component analysis for multi-subject fMRI data. NeuroImage. 2009 fMRI Course 3 2009 fMRI Course Approach 1 Approach • Separate ICA analysis for each subject [V. D. Calhoun, T. [V. Adali, V. McGinty, J. J. Pekar, T. Watson, and G. D. Pearlson, "FMRI Activation In A Visual"FMRI VisualPerception Task: Network Of Areas Detected Using The General Linear Model And Independent Linear Components Analysis," NeuroImage, vol. 14, pp. 1080-1088, 2001.] NeuroImage 1080- Example Example Press buttons (1-4) to (1indicate choice • Must select which components to compare between the individuals Sub 1 ICA ICA Sub N 1 2 3 4 ? 15 “events” events” … 0 15.4 31.5 47.0 300 Time (seconds) 2009 fMRI Course 5 2009 fMRI Course 6 Su bs Sub N Subject 1 Subject 1 4 1 Methods Methods • Scan Parameters • • • • • • • Single-shot EPI FOV = 24cm, 64x64 TR=1s, TE=40ms 18 slices Slice thickness = 5mm Gap = .5mm 300 volumes acquired SPM Results SPM Timing correction Motion correction Normalization Smoothing N=10 P<0.05 corrected • Preprocessing • • • • • ICA • An ICA estimation was performed on each of the ten subjects • Data were first reduced from 300 to 25 using principal component analysis (PCA) • Maps from each subject were inspected for similarity and similar maps were averaged across subjects (e.g. all visual areas were averaged together) • Group averaged maps were then thresholded at Z<3.1, colorized, and overlaid onto an EPI scan for visualization 2009 fMRI Course 7 SPM revealed a large network of areas including: •frontal eye fields •supplementary motor areas •primary visual •visual association •basal ganglia •thalamic, and an •(unexpectedly) large cerebellar activation •bilateral inferior parietal regions were deactivated (not deactivated shown) 2009 fMRI Course 8 ICA Results ICA ICA revealed a large network of similar areas including: •frontal eye fields (blue) •supplementary motor areas (green w/ outline) •primary visual (red) •visual association (red) •thalamic (red) •basal ganglia (green w/ outline) •a large cerebellar activation (red) •bilateral inferior parietal deactivations (not shown) deactivations ICA: Single Subject ICA: The ICA maps from one subject for the visual and basal ganglia components are depicted along with their time courses (basal ganglia in green and visual in pink) N=10 Z>3.1 ICA also revealed areas not identified by SPM including: •primary motor (green) •frontal regions anterior to the frontal eye fields (blue) •superior parietal regions (blue) 2009 fMRI Course 9 2009 fMRI Course Note that the visual time course precedes the motor time course 10 Event-Averaged Time Courses Event•Time courses from selected voxels in the raw data (a) and time courses produced by the ICA method (b). •In all cases the time courses are event-averaged (according eventto when the figure was presented) within each participant and then averaged across all ten participants. •Voxels from the raw data were selected by choosing a local maximum in the activation map and averaging the two surrounding voxels in each direction. •Dashed lines indicate the standard error of the mean. • • Approach 2 Approach • Group ICA (stacking images) [V. D. Calhoun, T. Adali, G. D. Pearlson, and J. J. Pekar, "A Method for Making Group Inferences Method From Functional MRI Data Using Independent Component Analysis," Hum. Brain Map., vol. 14, pp. Hum. 140-151, 2001.] 140[V. J. Schmithorst and S. K. Holland, "Comparison of Three Methods for Generating Group Statistical Inferences From Independent Component Analysis of Functional Magnetic Resonance Functional Imaging Data," J. Magn Reson. Imaging, vol. 19, pp. 365-368, 2004.] J. Reson. 365- • Components and time courses can be directly compared Sub 1 ICA Sub N Sub 1 Sub N 2009 fMRI Course 11 2009 fMRI Course 12 2 Group ICA Group Data ICA Ai Backreconstruction and Hypothesis Testing Backreconstruction Back-reconstruction Back1 • Single subject maps can be calculated by backbackreconstructing from the ICA analysis of all the subjects Si Subject 1 X Subject N A1 Subject i A AN S_agg F1† Y1 G1 AS ˆˆ † FM YM G M ˆ ˆ S i G i1A Fi† Yi single-subject map single- † ˆ s1i hypothesis test for component 1 (first row of Ŝi) ˆ Fi G i A single-subject singletime course full decomposition • These maps can then be tested for a significant amplitude by using a voxel-by-voxel t-test on the voxel- bysingle subject maps 2009 fMRI Course 13 2009 fMRI Course 14 Simulation Simulation Are the data separable? (Simulation) Are •A natural concern is whether the back-reconstructed maps backfrom individual subjects will be influenced by the other subjects in the group analysis •This simulation was performed in which one of the nine “subjects” had a structured, subjects” source #2 map (whereas all of the nine “subjects” had a subjects” similar, source #1 map). •As one can see, in this example, the back-reconstructed backICA maps are very close to the individual maps and there appears to be little to no influence between subjects Nine simulated source maps and time courses were generated, followed by an ICA estimation. The red lines indicate the t<4.5 boundaries 2009 fMRI Course 15 2009 fMRI Course 16 The Stationarity Assumption The Stationary source S common to all five “subjects” subjects” Sources S1-S5 S1differing across the five “subjects” subjects” Methods Methods • Scan Parameters • • • • • • • • • 9 slice Single-shot EPI FOV = 24cm, 64x64 TR=1s, TE=40ms Thickness = 5/.5 mm 360 volumes acquired Timing correction Motion correction Normalization Smoothing S S1 S2 S3 S4 S5 ICA results 2009 fMRI Course source #1 source #2 •The ICA estimation requires the data to be stationary across subjects •Some signals in the data (e.g. physiologic noise) will most likely *not* be stationary •However it is reasonable to assume the signal of interest (fMRI activation) will be stationary •A simulation was performed to examine how non-stationary nonsources would affect the results •One stationary signal (fMRI activation) and one non-stationary nonsignal were simulated for a fivefivesubject analysis •The ICA results reveal that the fMRI activation is preserved 17 Right • Preprocessing Left t (secs) 0 90 180 270 360 • ICA • An ICA estimation was performed on each of the nine subjects • Data were first reduced from 360 to 25 using PCA, the data were concatenated and reduced a second time from 225 to 20 using PCA • An ICA estimation was performed after which single subject maps and time courses were calculated • Group averaged maps were thresholded at t<4.5, colorized, and overlaid onto an EPI scan for visualization 2009 fMRI Course 18 3 Are the data separable? (fMRI experiment) Are Comparison with GLM Approach Comparison R L •The same slice from nine subjects when the right (red) and left (blue) visual fields were stimulated, (a) analyzed via linear modeling (LM), (b) back-reconstructed from a backgroup ICA analysis, or (c) calculated from an ICA analysis performed on each subject performed separately. A transiently task-related component is depicted in green. task•The results between the two ICA methods appear quite similar and match well with the LM results as well (note that there may be small differences due to different initial conditions for the ICA estimation) 2009 fMRI Course 19 2009 fMRI Course V.D. Calhoun, T. Adali, G.D. Pearlson, and J.J. Pekar, "A Method for Making Group Inferences From Functional MRI Data Using Independent Component Analysis," Hum. Brain Map., vol. 14, pp. 140-151, 2001. 20 Sorting/Calibrating Sorting/Calibrating • A ‘second-level’ or group analysis involves taking certain parameters (estimated by second-level’ ICA) such as the amplitude fit for fMRI regression models, or voxel weights, and voxel testing these within a standard GLM hypothesis-testing framework hypothesis- Example of spatial sorting Example Comp# R2 Subject Reg1 Reg2 1 10 4 2009 fMRI Course 0.81 0.81 0.017 1 2 1 2 1 2 1.89 2.28 0.28 0.65 -0.19 -0.10 0.02 0.66 2.19 2.03 -0.40 0.08 21 2009 fMRI Course 22 Example 1: ‘Default Mode’ Mask Example Mode’ • Using wfu pickatlas to define mask using regions reported in Rachle 2001 paper • • • • • Posterior parietal cortex BA7 Occipitoparietal junction BA 39 Precuneus Posterior cingulate Frontal Pole BA 10 ICA to identify ‘Default Mode’ Network ICA Mode’ Healthy Schizo • Smooth in SPM with same kernel used on fMRI data • Sort in GIFT using spatial sorting A.Garrity, G.D.Pearlson, K.McKiernan, D.Lloyd, K.A.Kiehl, and V.D.Calhoun, "Aberrant 'Default Mode' Functional Connectivity in Schizophrenia," to appear Am. J. Psychiatry, 2006. 2009 fMRI Course Healthy-Schizo (N=26/26) Healthy- A.Garrity, G.D.Pearlson, K.McKiernan, D.Lloyd, K.A.Kiehl, and V.D.Calhoun, "Aberrant 'Default Mode' Functional Connectivity in Schizophrenia," to appear Am. J. Psychiatry, 2006. 23 2009 fMRI Course Student: Abbie Garrity, current undergraduate student (neuroscience) Student: Abbie Garrity, current undergraduate student (neuroscience) 24 4 Spatial Sorting: Example 2 Spatial • Classification of Schizophrenia • Mapping the brain via intrinsic connectivity • • • • • Temporal Lobe Synchrony Temporal • Supervised Classification Step 1: Select Training Group Step 2: Use ICA to extract temporal lobe maps Step 3: Compute within-group mean images withinStep 4: Subtract the mean images Step 5: Set a positive and negative threshold Patients … Controls ICA … Sz1 SzN HC1 HCN t t 2009 fMRI Course 25 2009 fMRI Course Calhoun VD, Kiehl KA, Liddle PF, Pearlson GD: “Aberrant Localization of Synchronous Hemodynamic Activity in Auditory Cortex Reliably Characterizes Schizophrenia”. Biol Psychiatry 2004; 55842-849 26 Temporal Lobe Synchrony in Schizophrenia Temporal Temporal Lobe Synchrony in Schizophrenia Temporal • Step 6: Form classification measure (average the values within each each boundary and subtract) DF t , t , i i , IM t i , IM t .* MSK HC|SZ • Step 7: Optimize group discrimination (using a sensible error metric) metric) t min Err t , t DF t , t , i 0 DF t , t , i 0 iSz iHc • Step 8: Apply classification to new data t 2009 fMRI Course 27 2009 fMRI Course Calhoun VD, Kiehl KA, Liddle PF, Pearlson GD: “Aberrant Localization of Synchronous Hemodynamic Activity in Auditory Cortex Reliably Characterizes Schizophrenia”. Biol Psychiatry 2004; 55842-849 28 Temporal Sorting: fBIRN SIRP Task Temporal • Methods • Subjects & Task • 28 subjects (14 HC/14 SZ) across two sites • Three runs of SIRP task preprocessed with SPM2 Component 1: Bilateral Frontal/Parietal Component • ICA Analysis • All data entered into group ICA analysis in GIFT • ICA time course and image reconstructed for each subject, session, and session, component • Images: sessions averaged together creating single image for each subject and each component • Time courses: SPM SIRP model regressed against ICA time course • Statistical Analysis: • Images: all subjects entered into voxelwise 1-sample t-test in SPM2 and tthresholded at t=4.5 • Time courses: Goodness of fit to SPM SIRP model computed, beta weights for weights load 1, 3, 5 entered into Group x Load ANOVA 2009 fMRI Course fBIRN Phase II Data: www.nbirn.net; NCRR (NIH), 5 MOI RR 000827 (2002-2006) and 1 U24 RR0219921 (2006 onwards) (2002- 29 2009 fMRI Course fBIRN Phase II Data: www.nbirn.net; NCRR (NIH), 5 MOI RR 000827 (2002-2006) and 1 U24 RR0219921 (2006 onwards) (2002- 30 5 Component 1: Event-related Average of ICA time course Component Event- Component 2: Right Frontal, Left Parietal, Post. Cing. Component Cing. 2009 fMRI Course 31 2009 fMRI Course fBIRN Phase II Data: www.nbirn.net; NCRR (NIH), 5 MOI RR 000827 (2002-2006) and 1 U24 RR0219921 (2006 onwards) (2002- 32 Component 3: Temporal Lobe Component Example 2: Simulated Driving Paradigm Example * Drive Watch 0 180 360 600 2009 fMRI Course fBIRN Phase II Data: www.nbirn.net; NCRR (NIH), 5 MOI RR 000827 (2002-2006) and 1 U24 RR0219921 (2006 onwards) (2002- 33 2009 fMRI Course 34 Previous Work Previous • Walter, 2001. Driving Baseline Simulated Driving Results Baseline N=12 Higher Order Visual/Motor: Increases during driving; less Increases during watching. Low Order Visual: Increases during driving; less during watching. Motor control: Increases only during driving. Vigilance: Decreases only during driving; amount proportional to speed. Error Monitoring & Inhibition: Decreases only during driving; rate proportional to speed. Visual Monitoring: Increases during epoch transitions. Watching “Our results suggest that simulated driving engages mainly areas concerned with perceptualmotor integration and does not engage areas associated with higher cognitive functions.” “our study suggests that the main ideas of cognitive psychology used in the design of cars, in the planning of respective behavioral experiments on driving, as well as in traffic related political decision making (i.e. laws on what drivers are supposed to do and not to do during driving) may be inadequate, as it suggests a general limited capacity model of the psyche of the driver which is not supported by our results. If driving deactivates rather than activates a number of brain regions the quests for the adequate design of the man-machine interface as well as for what the driver should and should not do during driving is still widely open.” 2009 fMRI Course 35 2009 fMRI Course * Drive Watch 36 V. D. Calhoun, J. J. Pekar, V. B. McGinty, T. Adali, T. D. Watson, and G. D. Pearlson, "Different Activation Watson, Dynamics in Multiple Neural Systems During Simulated Driving," Hum. Brain Map., vol. 16, pp. 158-167, 2002. Hum. 158- 6 SPM Results SPM Interpretation of Results Interpretation 2009 fMRI Course Calhoun, V. D., Pekar, J. J., and Pearlson, G. D. “Alcohol Intoxication Effects on Simulated Driving: Exploring AlcoholDose Effects on Brain Activation Using Functional MRI”. Neuropsychopharmacology 2004. 37 2009 fMRI Course 38 Functional Network Connectivity Functional (between groups) Key: : ρpatient > ρcontrol : ρcontrol > ρpatient G: Temporal A: Default B: Parietal FNC Software FNC F: Frontal C: L. & M. Visual Cortical Areas E: Frontal Parietal Subcortical D: Frontal Temporal Parietal 2009 fMRI Course 39 2009 fMRI Course 40 Demo Demo • 3 subject ICA • Sorting • Component Explorer (split time courses, event-related eventaverage) • Orthogonal Viewer • Composite Viewer • Examine Regression Parameters • Taking Images/Timecourses from GIFT to SPM Images/Timecourses 2009 fMRI Course 41 7 ...
View Full Document

This note was uploaded on 02/10/2010 for the course TBE 2300 taught by Professor Cudeback during the Spring '10 term at Webber.

Ask a homework question - tutors are online