James Wells Catherine Koene Pamela Feldkamp Josh Proksch Lorcan Duffy Francis Moniz.
Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$...
Transcript of Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$...
![Page 1: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/1.jpg)
Why ‘n How: Func.onal connec.vity MRI
Koene Van Dijk
Mar.nos Center for Biomedical Imaging, Febr 21, 2013
![Page 2: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/2.jpg)
Why
![Page 3: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/3.jpg)
Spontaneous Activity Within Seed Region!
Percent signal m
odula.
on
-‐40
-‐30
-‐20
-‐10
0
10
20
30
40
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
Biswal et al., MRM, 1995
How
![Page 4: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/4.jpg)
SEED BASED ICA
MOTO
R VISU
AL
DEFAU
LT
ATTENTION
Van Dijk et al., 2010 Yeo et al., 2011
![Page 5: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/5.jpg)
Func.onal connec.vity MRI (fcMRI)
Measures temporal coherence among brain regions (Biswal et al., 1995)
Shows exis.ng func.onal-‐anatomical networks.
Most networks show coherent ac.vity both during rest and during tasks (Fransson, 2006)
Yeo et al., 2011
![Page 6: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/6.jpg)
Cerebro-‐Cerebellar Circuits
Krienen and Buckner, 2009, CerebCortex
![Page 7: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/7.jpg)
Cerebro-‐Cerebellar Circuits
Lu et al., 2011, JNeurosci
![Page 8: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/8.jpg)
fcMRI methods
1. Seed based analysis
2. Independent component analysis (ICA)
3. Graph analy.c approaches
4. Clustering
![Page 9: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/9.jpg)
Seed based analysis • Standard fMRI pre-‐processing:
– Slice .me correla.on
– Mo.on correc.on
– Spa.al normaliza.on to common atlas space (Talairach / MNI)
– Spa.al smoothing (Gaussian filter)
• Addi.onal pre-‐processing:
– Bandpass filter (retaining frequencies slower than 0.08 Hz or between 0.009 – 0.08 Hz)
– Removal of any residual effects of mo.on by regressing out the mo.on correc.on parameters
– Removal of physiological noise
(i.e. mostly concerned with variability in heart rate and respira.on)
![Page 10: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/10.jpg)
Removal of physiological noise • By regressing out signals in the brain that are believed to contain noise:
– Signal from ventricles
– Signal from white maier
– Signal averaged over the whole brain (Desjardins et al., 2001; Corfield et al., 2001; Macey et al. 2004)
(Dagli et al. 1999; Windischberger et al. 2002)
• By measuring variability in heart rate and respira.on and regressing it out:
– RETROICOR (Glover et al., 2002)
– RVT (Birn et al., 2006)
– RVHRCOR (Chang and Glover, 2009)
• By es.ma.ng variability in heart rate and respira.on and regressing it out:
– PESTICA (Beall and Lowe, 2007; 2010)
– CompCor (Behzadi et al., 2007; also implemented in Sue Whikield-‐Gabrieli’s Conn toolbox)
– CORSICA (Perlbarg et al., 2007)
![Page 11: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/11.jpg)
Biswal et al., 1995 Wise et al., 2004
Carbon dioxide fluctua.ons Low frequency BOLD fluctua.ons
Removal of physiological noise
![Page 12: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/12.jpg)
Ventricles and white maier signal
Ventricles
White maier
Background
Grey maier
Windischberger et al. 2002
![Page 13: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/13.jpg)
Whole brain signal
Corfield et al., 2001
Whole brain signal is related to end-‐.dal carbondioxide during periods of hypercapnia
![Page 14: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/14.jpg)
Removal of whole brain signal Pros:
• Captures breathing varia.ons (Corfield et al., 2001)
• Known to remove noise (Macy et al., 2004)
• Increased specificity of func.onal connec.vity maps of the motor system (Weisenbacher et al., 2009)
• Shows fine neuroanatomical specificity not seen without global signal correc.on (Fox et al., 2009)
![Page 15: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/15.jpg)
Removal of whole brain signal Cons:
See for discussion: Vincent et al., 2006; Fox et al. 2009; Murphy et al. 2009; Weissenbacher et al., 2009; Van Dijk et al., 2010; Saad et al., 2012
Murphy et al. 2009; Van Dijk et al., 2010
Before Aner
![Page 16: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/16.jpg)
Removal of whole brain signal What does this prac.cally mean?
• Whenever possible collect heart rate (with e.g. pulse oximeter or ECG) and respira.on signals (with e.g. pneuma.c belt around the abdomen or with a mouth piece that measures airflow)
• Removal of the whole brain signal is a viable step especially if one has not measured the actual variability in heart rate and respira.on.
• Nega.ve correla.ons aner whole-‐brain signal regression should be interpreted with utmost cau.on.
See for discussion: Vincent et al., 2006; Fox et al. 2009; Murphy et al. 2009; Weissenbacher et al., 2009; Van Dijk et al., 2010; Saad et al., 2012
![Page 17: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/17.jpg)
Basic measures seed based analysis
B. A map from a seed region of interest:
A. Correla.on values between regions of interest:
![Page 18: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/18.jpg)
fcMRI methods
1. Seed based analysis
2. Independent component analysis (ICA)
3. Graph analy.c approaches
4. Clustering
![Page 19: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/19.jpg)
Independent component analysis (ICA)
Data is decomposed into a set of spa.ally independent maps and corresponding .me courses
SINGLE SUBJEC
T
MELODIC: hip://fsl.fmrib.ox.ac.uk/fsl/melodic/ Beckmann et al., 2005
GIFT: hip://mialab.mrn.org/sonware/ Calhoun et al., 2001
![Page 20: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/20.jpg)
Independent component analysis (ICA) SINGLE SUBJEC
T MULTI SUBJEC
T
MELODIC: hip://fsl.fmrib.ox.ac.uk/fsl/melodic/ Beckmann et al., 2005
GIFT: hip://mialab.mrn.org/sonware/ Calhoun et al., 2001
![Page 21: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/21.jpg)
Zou et al., 2010
![Page 22: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/22.jpg)
Noise components
Beckmann et al., 2000 & hip://www.fmrib.ox.ac.uk/fslcourse/lectures/melodic.pdf
![Page 23: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/23.jpg)
Neuronal network components
Damoiseaux et al., 2006
![Page 24: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/24.jpg)
1) Concat ICA: Mul.ple fMRI data sets are concatenated temporally and ICA is applied in order to iden.fy large-‐scale pa+erns of func0onal connec0vity in the sample
2) Dual regression: This is used to iden.fy, within each of the individual subjects’ fMRI data, spa.al maps and associated .mecourses corresponding to the mul.-‐subject ICA components.
First papers using dual regression approach: -‐ Filippini, et al., Dis.nct paierns of brain ac.vity in young carriers of the APOE-‐epsilon4 allele. Proc Natl Acad Sci U S A. 2009;106(17):7209-‐14. -‐ Glahn et al., Gene.c control over the res.ng brain. Proc Natl Acad Sci U S A. 2010;107(3):1223-‐8.
Beckmann et al., poster HBM, 2009
![Page 25: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/25.jpg)
Seed based Pros:
• Quan.fica.on of network strength: one number per subject (easy to relate to other measures such as behavior, white maier (Andrews-‐Hanna et al., 2007), amyloid pathology (Hedden et al., 2009))
• Input for other other metrics: graph analy.c approaches, clustering, mul.variate approaches
Cons:
• One needs to choose networks based on prior knowledge of brain systems (e.g. from anatomy, seed regions from the literature)
• Methods for removal of physiological noise not without controversy (Fox et al. 2009; Murphy et al. 2009)
ICA Pros:
• No prior knowledge of brain systems necessary
• Automa.c removal of physiological noise (to a certain extent)
Cons:
• Obtaining one number per subject is not straighkorward (goodness-‐of-‐fit metric is not without controversy (Franco et al., 2009) )
![Page 26: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/26.jpg)
More regarding methods
![Page 27: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/27.jpg)
hip://jn.physiology.org/content/103/1/297.full Free online access:
Test-‐retest reliability
Effects of scan length
Seed-‐based analysis versus ICA
“Time on task” effects within a res.ng res.ng-‐state run
Differences between:
Two scans of 6 min versus one scan of 2 min
Temporal resolu.on: TR=2.5 versus TR=5.0
Spa.al resolu.on: 2x2x2 versus 3x3x3
Effects of task/instruc.on (eyes open, eyes closed, word classifica.on task)
![Page 28: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/28.jpg)
Reliability of network measures
SUBJECT 1 SUBJECT 2 SUBJECT 3 SUBJECT 4 SUBJECT 5 SUBJECT 6
SESSION 1
SESSION 2
OVER
LAP
Van Dijk et al., 2010, JNeurophys
![Page 29: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/29.jpg)
SUBJECT 1 SUBJECT 2 SUBJECT 3 SUBJECT 4 SUBJECT 5 SUBJECT 6
SESSION 1
SESSION 2
OVER
LAP
Reliability of network measures
Van Dijk et al., 2010, JNeurophys
![Page 30: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/30.jpg)
Short acquisi.on .mes are sufficient
1 minute 3 minutes 6 minutes 12 minutes
Van Dijk et al., 2010, JNeurophys
![Page 31: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/31.jpg)
Signal
Noise
Short acquisi.on .mes are sufficient
Run length (min)
Van Dijk et al., 2010, JNeurophys
![Page 32: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/32.jpg)
Confounding effects of head mo.on
![Page 33: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/33.jpg)
![Page 34: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/34.jpg)
Transla.onal mo.on of 3 subjects Mo.
on (m
m)
0
0.4
0.8
1.2
1.6
1 51 101 151 201
Number of volumes
Van Dijk et al., 2012, Neuroimage
![Page 35: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/35.jpg)
Head mo.on is a problem
Decile 5 > Decile 6
r(z)
0.20
0.08
Mean Mo.on (mm)
Freq
uency (%
)
Van Dijk et al., 2012, Neuroimage
![Page 36: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/36.jpg)
Mo.on decreases correla.on strength in large-‐scale networks
r = -‐0.18 r = -‐0.16
Van Dijk et al., 2012, Neuroimage
![Page 37: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/37.jpg)
Mo.on increases local func.onal coupling
rn = 0.11 rn = 0.15
Van Dijk et al., 2012, Neuroimage
![Page 38: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/38.jpg)
3. Carefully consider effects of head mo.on in studies that contrast groups: – Children vs adults
– Old vs young
– Pa.ents vs controls
4. Improve effec.veness of current regression techniques
5. Consider “scrubbing” epochs were mo.on occurred (e.g. Power et al., 2012)
1. Head mo.on has a significant effect on measures of network strength
2. Most variance in fcMRI metrics is not related to mo.on
Conclusions regarding head mo.on
Van Dijk et al., 2012, Neuroimage
![Page 39: Why‘n$How:$$ Func.onal$connec.vity$MRInmr.mgh.harvard.edu/whynhow/fcMRI_vanDijk.pdfWhy‘n$How:$$ Func.onal$connec.vity$MRI! Koene$Van$Dijk$ Mar.nos$Center$for$Biomedical$Imaging,$Febr21,2013](https://reader030.fdocuments.in/reader030/viewer/2022040818/5e628324dda77805fa086c89/html5/thumbnails/39.jpg)
E-‐mail: [email protected]
Web: h+p://www.nmr.mgh.harvard.edu/~kdijk/
References:
Van Dijk KRA, Hedden T, Venkataraman A, Evans KC, Lazar SW, and Buckner RL (2010) Intrinsic Func.onal Connec.vity As a Tool For Human Connectomics: Theory, Proper.es, and Op.miza.on. Journal of Neurophysiology. 103: 297-‐321. Full text at: hip://jn.physiology.org/content/103/1/297.full
Van Dijk KRA, Sabuncu MR, and Buckner RL. (2012) The Influence of Head Mo.on on Intrinsic Func.onal Connec.vity MRI. NeuroImage. 59(1):431-‐8. Abstract at: hip://www.sciencedirect.com/science/ar.cle/pii/S1053811911008214 (send me an e-‐mail if you would like a PDF)
Thank you for your aien.on!