Damoiseaux JS, Rombouts SA, Barkhof F, Scheltens P, Stam CJ, Smith SM, Beckmann CF.
Consistent resting-state networks across healthy subjects.
Proc Natl Acad Sci U S A. 2006 Sep 12;103(37):13848-53.
Please login to recommend the paper.
To make a comment you must login or register.
In a landmark functional imaging paper (1), Marc Raichle described a network of brain regions, featuring the posterior cingulate cortex, medial prefrontal cortex, and temporoparietal junctions, whose activity decreased during the performance of externally cued tasks and increased during task-free periods across a wide variety of cognitive tasks. He hypothesized that this set of regions constituted a "default mode" of brain function, a network of regions tonically active at rest but attenuated during the performance of externally cued cognitive tasks. Subsequent work using resting-state functional connectivity MRI (2-4) has confirmed that these regions exhibit strong temporal correlations in spontaneous fMRI signal fluctuations during task-free, "resting-state" scans. Alzheimer disease (AD) researchers have shown interest in this default mode network because of its remarkable overlap with those brain regions (posterior cingulate and temporoparietal junctions in particular) that typically show reduced glucose metabolism in patients with AD. A series of recent fMRI studies have shown group-level differences in this network between AD patients and healthy controls (5-7). A multimodal study by Buckner and colleagues (8) has suggested that all roads lead to the default mode network. That is, across the various imaging modalities used to study AD—functional MRI, longitudinal structural MRI, flurodeoxyglucose PET, and PET with the amyloid tagging Pittsburgh-B compound —all implicate core regions in the default mode network, notably the posterior cingulate cortex, the temporoparietal junctions, and, more variably, the medial prefrontal cortex and hippocampus.
On this background, the new paper by Damoiseaux and colleagues—a combined effort between imaging labs in Leiden and Oxford—offers some important insights into the strength and consistency of activity in the default mode network. These new insights, in turn, have important implications for the use of resting-state fMRI in the study of AD. The authors used independent component analysis (ICA) to isolate and characterize the default mode network (and several other canonical neural networks) across 10 healthy subjects scanned for 9.5 minutes on a 1.5T magnet. Subjects were instructed simply to “lie with their eyes closed, think of nothing in particular, and not fall asleep.” Each subject was scanned twice with roughly 1 week between scans. Using a bootstrapping approach, they show that the default mode network is remarkably consistent across both subjects and repeated sessions. Furthermore, they have expanded on the typical ICA approach, limited to information on the temporal coherence of resting-state fluctuations between brain regions, to include information on the amplitude of these fluctuations. They estimate the amplitude of these fluctuations in the default mode network to be between 2 and 3 percent in core regions such as the posterior cingulate and lateral parietal lobes. This is on a par with the 2-3 percent signal change detected in typical task-activation fMRI paradigms.
Resting-state fMRI has several advantages over standard task-activation fMRI when considering functional biomarker candidates in AD. It is difficult, even at the mild stages of AD, to get cognitively impaired subjects to perform a cognitive task in the scanner. Differences in performance between AD subjects and healthy controls represent a thorny confound when interpreting activation differences. Task-activation fMRI protocols are generally cumbersome to perform, requiring extra hardware to project stimuli and record responses—one fact among several that has restricted its use, almost exclusively, to academic hospital settings. Resting-state fMRI, by contrast, is something that subjects with relatively advanced dementia can still undergo. It does not have performance confounds, it could be run by an MRI technician in a community hospital, and it can be quantified at the single-subject level in an automated fashion using ICA (6). The findings reported by Damoiseaux et al. provide encouraging evidence that resting-state fMRI signal in the default mode network is sufficiently strong and consistent to merit wider testing as a biomarker in AD. The important addition of amplitude information to the temporal coherence metric used previously should help enhance the sensitivity and specificity of this approach substantially.
To close, in the spirit of freeware and shared data, I would point out that 1) the MELODIC ICA software used in this study can be downloaded for free from the Oxford group (it does not yet include all the tools used in the current study, but this is expected in the next release) and 2) there is a superb collection of near-rest fMRI data in AD subjects and controls available at the (go to fMRI Datacenter, request the dataset submitted by Randy Buckner, John Morris, and the group at Washington University in St. Louis, and it will be sent to you).
Raichle ME, MacLeod AM, Snyder AZ, Powers WJ, Gusnard DA, Shulman GL.
A default mode of brain function.
Proc Natl Acad Sci U S A. 2001 Jan 16;98(2):676-82.
Greicius MD, Krasnow B, Reiss AL, Menon V.
Functional connectivity in the resting brain: a network analysis of the default mode hypothesis.
Proc Natl Acad Sci U S A. 2003 Jan 7;100(1):253-8.
Fox MD, Snyder AZ, Vincent JL, Corbetta M, Van Essen DC, Raichle ME.
The human brain is intrinsically organized into dynamic, anticorrelated functional networks.
Proc Natl Acad Sci U S A. 2005 Jul 5;102(27):9673-8. Epub 2005 Jun 23
Spontaneous low-frequency BOLD signal fluctuations: an fMRI investigation of the resting-state default mode of brain function hypothesis.
Hum Brain Mapp. 2005 Sep;26(1):15-29.
Lustig C, Snyder AZ, Bhakta M, O'Brien KC, McAvoy M, Raichle ME, Morris JC, Buckner RL.
Functional deactivations: change with age and dementia of the Alzheimer type.
Proc Natl Acad Sci U S A. 2003 Nov 25;100(24):14504-9.
Greicius MD, Srivastava G, Reiss AL, Menon V.
Default-mode network activity distinguishes Alzheimer's disease from healthy aging: evidence from functional MRI.
Proc Natl Acad Sci U S A. 2004 Mar 30;101(13):4637-42.
Rombouts SA, Barkhof F, Goekoop R, Stam CJ, Scheltens P.
Altered resting state networks in mild cognitive impairment and mild Alzheimer's disease: an fMRI study.
Hum Brain Mapp. 2005 Dec;26(4):231-9.
Buckner RL, Snyder AZ, Shannon BJ, LaRossa G, Sachs R, Fotenos AF, Sheline YI, Klunk WE, Mathis CA, Morris JC, Mintun MA.
Molecular, structural, and functional characterization of Alzheimer's disease: evidence for a relationship between default activity, amyloid, and memory.
J Neurosci. 2005 Aug 24;25(34):7709-17.