Summary: Zooming out to image larger areas of the brain while using fMRI technology allows researchers to capture additional relevant information, providing a better understanding of neural interaction.
Researchers have learned a lot about the human brain using functional magnetic resonance imaging (fMRI), a technique that can provide insight into brain function. But typical fMRI methods can miss key information and provide only part of the picture, according to Yale researchers.
In a new study, they evaluated various approaches and found that zooming out and widening the field of view capture additional relevant information that narrow focusing misses, providing better understanding of neural interaction.
Additionally, these more encompassing results may help address the problem of neuroimaging reproducibility, in which some results presented in studies cannot be replicated by other researchers.
The results were published on August 4 in Proceedings of the National Academy of Sciences.
Studies using fMRI generally focus on small areas of the brain. As an example of this approach, researchers look for regions of the brain that become more “active” when a particular activity is performed, focusing on small areas with the strongest activation. But a growing body of evidence shows that brain processes, and complex processes in particular, are not confined to small parts of the brain.
“The brain is a network. It’s complex,” said Dustin Scheinost, associate professor of radiology and biomedical imaging and lead author of the study. Oversimplifying, he said, leads to inaccurate conclusions.
“For more sophisticated cognitive processes, it’s unlikely that many areas of the brain are completely uninvolved,” added Stephanie Noble, postdoctoral associate in Scheinost’s lab at Yale School of Medicine and lead author of the study.
Focusing on small areas leaves out other regions that may be involved in the behavior or process under study, which may also affect the direction of future research.
“You develop this incorrect picture of what’s really going on in the brain,” she said.
For the study, researchers assessed how well fMRI scans across a range of scales were able to detect effects or changes in fMRI signals when participants perform different activities, revealing which parts of the brain are engaged.
They used data from the Human Connectome Project, which collected brain scans of individuals performing different tasks related to complex processes such as emotion, language and social interactions.
The research team looked for effects in very small parts of the brain network, such as connections between just two areas, as well as in clusters of connections, extended networks, and whole brains.
They found that the larger the scale, the better they were able to detect effects. This ability to detect effects is known as “potency”.
“We get better potency with these larger-scale methods,” Noble said.
At the smallest scales, the researchers were only able to detect about 10% of the effects. But at the network level, they could detect more than 80%.
The trade-off for the extra power was that the larger views did not relay as spatially accurate information as the smaller-scale scans. For example, at the smallest scale, the researchers could say with confidence that the effects they observed were occurring throughout the small area.
At the network level, however, they could only say that the effects were happening over a large part of the network, without identifying exactly where in the network.
The goal, says Noble, is to balance the pros and cons of the different methods.
“Do you prefer to be very confident about a small part of the relevant information, in other words, to have a very clear picture of the tip of the iceberg?” she says.
“Or do you prefer to have a very large picture of the whole iceberg that is maybe a bit blurry but gives you a sense of the complexity and the large spatial scale of where things are happening in the brain?
For other researchers, this approach is simple to implement, and Noble said she looks forward to seeing how other scientists use it.
She notes that the fields of psychology and neuroscience, including neuroimaging, have faced a problem of reproducibility. And the low power of fMRI scans contributes to this: low-powered studies tell only small parts of the story, which can be seen as contradictory rather than parts of a whole.
Increasing the power of fMRI, as she and her colleagues have done here by scaling up their analyses, could be a way to address reproducibility challenges by exposing how seemingly conflicting results can actually be. be harmonious.
“Moving up the food chain, so to speak, going from a very low level to more complex networks buys you a lot more power,” Scheinost said. “It’s one of the tools we can use to solve the reproducibility problem.”
And scientists shouldn’t throw the baby out with the bathwater, Noble said. There is a lot of good work going on to improve methods and build rigor, and fMRI is still a valuable tool, she said: “I think the assessment of power, rigor and reproducibility is healthy for any field. Especially the one that deals with the complexity of living beings and mental processes.
Noble is currently developing a “power calculator” for fMRI, to help others design studies to achieve the desired level of power.
About this neuroimaging research news
Author: Mallory Locklear
Contact: Mallory Locklear–Yale
Image: Image is in public domain
Original research: Free access.
“Improving the Power of Functional Magnetic Resonance Imaging by Going Beyond Cluster-Level Inference” by Stephanie Noble et al. PNAS
Improving the power of functional magnetic resonance imaging by going beyond cluster-level inference
Inference in neuroimaging typically occurs at focal brain areas or circuits. Yet increasingly powerful studies paint a much richer picture of large-scale effects distributed throughout the brain, suggesting that many focal reports may only reflect the tip of the iceberg of sub-brain effects. underlyings.
How focal and large-scale perspectives influence the inferences we make has yet to be fully assessed using real data.
Here, we compare sensitivity and specificity between procedures representing multiple levels of inference using an empirical benchmarking procedure that resamples task-based connectomes from the Human Project dataset. Connectome (∼1000 subjects, 7 tasks, 3 resampling group sizes, 7 inferential procedures).
Only the large-scale (network and whole-brain) procedures achieved the traditional 80% statistical power level to detect a medium effect, reflecting >20% more statistical power than the focal (edge and cluster) procedures. There was also a significant increase in power for the false discovery rate – compared to procedures controlling the error rate per family.
The downsides are quite limited; the loss of specificity for the large-scale and FDR procedures was relatively modest compared to the gains in power. Additionally, the large-scale methods we introduce are simple, fast, and easy to use, providing researchers with a direct starting point.
It also points to the promise of more sophisticated methods at scale not only for functional connectivity, but also for related areas, including task-based activation.
Overall, this work demonstrates that changing the scale of inference and choosing the FDR control are both immediately feasible and can help address the statistical power issues that plague typical studies in the field.