Overview: By zooming out to image larger areas of the brain while using fMRI technology, researchers can capture additional relevant information, providing a better understanding of the neural interaction.
Researchers have learned a lot about the human brain through functional magnetic resonance imaging (fMRI), a technique that can provide insight into brain function. But typical fMRI methods may miss important information and provide only part of the picture, Yale researchers say.
In a new study, they evaluated different approaches and found that zooming out and taking a wider field of view captures additional relevant information that omits narrow focus, providing a better understanding of neural interaction.
Furthermore, these more comprehensive results may help address the reproducibility problem of neuroimaging, where some findings presented in studies cannot be reproduced by other researchers.
The findings were published Aug. 4 in Proceedings of the National Academy of Sciences.
Studies with fMRI usually focus on small areas of the brain. As an example of this approach, researchers look for areas of the brain that become more “active” when a particular activity is performed, targeting small areas of strongest activation. But a growing body of evidence shows that brain processes, especially complex ones, are not limited 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 senior author of the study. Oversimplification, he said, leads to inaccurate conclusions.
“For more advanced cognitive processes, many areas of the brain are unlikely to be involved at all,” added Stephanie Noble, a postdoctoral associate in Scheinost’s lab at Yale School of Medicine and lead author of the study.
Focusing on small areas excludes other regions that may be involved in the behavior or process being studied, which may also influence the direction of future research.
“You develop this inaccurate picture of what’s really going on in the brain,” she said.
For the study, researchers assessed how well fMRI analyzes at different scales are able to detect effects or changes in fMRI signals as participants perform different activities, revealing which parts of the brain are involved.
They used data from the Human Connectome Project, which collected brain scans of individuals as they performed various 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 clusters of connections, widespread networks and whole brains.
They found that the larger the scale, the better they could detect effects. This ability to detect effects is known as ‘power’.
“We’re getting more power with these broader methods,” Noble said.
On the smallest scale, researchers were only able to detect about 10% of the effects. But at the network level, they were able to detect more than 80% of them.
The trade-off for the extra power was that the wider views didn’t relay information as spatially exact as that of smaller-scale analyses. For example, on the smallest scale, researchers could confidently say that the effects they observed occurred in the very small area.
However, at the network level, they could only say that the effects occurred in a large part of the network, not exactly where in the network.
The goal, Noble says, is to weigh up the pros and cons of the different methods.
“Would you rather have a lot of confidence in a small portion of the relevant information, in other words, have a very clear view of just the tip of the iceberg?” she said.
“Or would you rather have a really big view of the whole iceberg that might be a little blurry, but gives you a sense of the complexity and broad spatial scale of where things take place in the brain?”
For other researchers, this approach is easy to implement, and Noble said she’s looking forward to seeing how other scientists use it.
She notes that the fields of psychology and neuroscience, including neuroimaging, have had a reproducibility problem. And low power in fMRI analyzes contributes to this: Low-power studies reveal only small parts of the story, which can be seen as contradictory rather than parts of a whole.
Increasing fMRI power, as she and her colleagues here did by increasing the scale of their analyses, could be one way to address reproducibility challenges by showing how seemingly contradictory results actually harmonize. can be
“By stepping up the food chain, so to speak, from a very low level to more complex networks, you get a lot more power,” Scheinost said. “This is one of the tools we can use to overcome the reproducibility problem.”
And scientists shouldn’t throw the baby out with the bathwater, Noble said. A lot of good work is being done to improve methods and increase accuracy, and fMRI is still a valuable tool, she said: “I think assessing strength, accuracy and reproducibility is healthy for any field. Especially one that focuses on deals with the complexity of living things and mental processes.”
Noble is now developing a “power calculator” for fMRI, to help others design studies in a way that achieves a desired power level.
About this neuroimaging research news
Author: Mallory Locklear
Contact: Mallory Locklear – Yalea
Image: The image is in the public domain
Original research: Open access.
“Improving power in functional magnetic resonance imaging by moving beyond cluster-level inference” by Stephanie Noble et al. PNAS
Improving capability in functional magnetic resonance imaging by going beyond cluster-level inference
Inference in neuroimaging usually occurs at the level of focal brain regions or circuits. But increasingly, well-researched studies provide a much richer picture of large-scale effects spanning the brain, suggesting that many focal reports only reflect the tip of the iceberg of underlying effects.
How focal versus broad perspectives affect the inferences we make has not yet been fully evaluated using real data.
Here we compare the sensitivity and specificity between procedures representing multiple levels of inference using an empirical benchmarking procedure resampling task-based connectomes from the Human Connectome Project data set (∼1,000 subjects, 7 tasks, 3 resampling group sizes, 7 inferential procedures).
Only large-scale (network and whole brain) procedures obtained the traditional 80% statistical power level to detect an average effect, which corresponds to >20% more statistical power than focal (edge and cluster) procedures. Power also increased significantly for false discoveries — compared to family-wide error checking procedures.
The drawbacks are quite limited; the loss of specificity for large scale and FDR procedures was relatively modest compared to the gain in strength. In addition, the large-scale methods we introduce are simple, fast and easy to use, providing a clear starting point for researchers.
This also points to the promise of more advanced, broad methods for not only functional connectivity, but also related fields, including task-based activation.
Overall, this work demonstrates that shifting the scale of inference and choosing FDR control are both immediately feasible and can help solve the statistical power problems that plague typical studies in the field.