Scientists have created the first reference charts for the human brain, mapping its growth from infancy to 100 years old. Now, they have to grapple with difficult ethical questions about how they should — and perhaps shouldn’t — be used.
The reference charts are visualizations created from aggregating analyses of over 120,000 brain scans to show ranges in brain size, or gray matter volume, for each age. They also track the human brain’s rapid expansion early in life and its gradual shrinking over time. The researchers primarily developed the charts to provide a standardized measurement that other neuroscientists could use for brain imaging research, with the hope that maybe one day it could lead to a tool used in clinics.
“It’s an absolutely spectacular advancement in neuroscience and neuroimaging,” said Judy Illes, professor of neurology and neuroethics at the University of British Columbia.
Right now, the tool is designed purely for research and comes with caveats, including that it’s limited by a lack of age and geographic diversity in the datasets used to build it.
As the researchers build on the charts and as they’re rolled out for use in scientific studies, there are three important questions experts say will have to be answered about how the charts can be used.
What do scientists mean by a benchmark?
The reference charts are designed to be used as benchmarks — or a population average of sorts — that are standardized and can be used across studies. With so many different neuroimaging studies, datasets, and imaging approaches, it’s a technical challenge to understand how the brain evolves across the lifespan, particularly when researchers aren’t able to easily image someone’s brain over time. By having a benchmark built on data from different studies, scientists running other research studies can all use a consistent measure for any time point in the lifespan, and compare it to population averages at the same time.
That kind of reference can help researchers study the structure of the human brain, traits that are specific to age, how the brain develops over time, and how different diseases can affect it. Unlike other benchmarks — like the height and weight percentiles used to measure a baby’s growth — the brain reference charts aren’t ready to be used in clinical care. But even in research, benchmarking can still suggest something universal, and plays into a complex conversation about the role of any sort of “average” in neuroscience research.
“There are always challenges when you aggregate or consolidate data. Invariably people who represent the tails become underrepresented,” said Illes.
Marcello Ienca, a research fellow at the department of health sciences and technology at the Swiss Federal Institute of Technology Zurich, said that an implicit concern is that developing any new benchmark of the human brain could risk medicalization of — and discrimination against — neurodiversity. Ienca said the charts should be used to advance science, but also open the door for a public discussion about the importance of neurodiversity within neuroscience research.
And when the charts are being used as a benchmark, experts said, they need to be carefully accompanied by context whenever they’re used. The study authors said that positions within it don’t confer any value judgment nor any specific clinical disposition. Each brain is unique, resilient, and compensates in different ways in response to damage or dysfunction — a phenomenon known as plasticity, and the reference charts are based on measures of brain structure, not function.
“Just because an individual falls somewhere on a trajectory that isn’t favorable compared to their peers, doesn’t mean it will be a bad outcome. There’s a lot of individual variability in growth trajectories — both for the brain and human development,” said Lucina Uddin, professor in the department of psychiatry and biobehavioral sciences at the University of California, Los Angeles.
How do you avoid creating or exacerbating stereotypes?
One critical question is whether people could make associations between brain charts and other features, including education, race and ethnicity, employment status, and criminality, particularly if brain charts are combined with other datasets. There’s a risk that such a use could perpetuate or create inaccurate racist or gender-based stereotypes or fuel discrimination against individuals.
It’s an issue that has also been raised with regard to genetics research. In genetics, such instances of discrimination have already been documented, including cases in which insurance companies and employers have used genetic data to deny hundreds of people employment or insurance based on genetic predisposition to illness.
“We shouldn’t look at this [tool] as the holy grail. We don’t want to get to a point where we say based on your brain chart, you’re not qualified for a job,” said Laura Cabrera, an associate professor of engineering science and mechanics and chair in neuroethics at Penn State University.
For now, experts say that’s not an imminent concern, because of how the researchers designed the scientific framework for the study. If the charts become used more widely, the tool needs to be used in a focused manner with consideration of the negative impacts on self-image that could happen to people who fall outside of normal ranges.
What risks could research carry?
Experts say that for all the research benefits the reference charts could bring, they also come with clear risks if they’re used in the wrong way. Right now, the charts are nowhere close to being used clinically — so many of the most troubling scenarios around clinical use aren’t a reality at this point. And in modern medicine and science, people can’t get their brain scanned for no clinical purpose or anything other than possible disease. But that could change in the long run, particularly given the growth of direct-to-consumer neurotechnologies.
And even the use of reference charts in research carries ethical concerns. Nita Farahany, professor of law and philosophy at Duke Law School, said the charts “holds the potential for important advancements in understanding changes in neurodevelopment over time” — but noted that it also raises the question of what happens if researchers compare brain features between different populations, continents, and cultures. The same questions could be raised by studies that examine associations between the charts and non-clinical outcomes.
Ultimately, “the use of the models — ethically, scientifically, or clinically — comes down to the people who use them. People with expertise also share a responsibility to flag unethical use if it occurs,” said Aaron Alexander-Bloch, director of the Brain-Gene-Development Laboratory at the Children’s Hospital of Philadelphia and senior author of the study.
Illes said it’s critical for neuroethicists to be actively working with neuroscientists to help them ask the right research questions and avoid inappropriate applications of technology. This could include developing an ethics guide that specifies how the tool could be applied in research, and specifying what are the risks of other sensitive research topics including longer-term associations with demographics not related to health and changes in people who don’t need brain scans.
“The goal is never to impede progress or fear-monger, but to very positively anticipate problems that may come down the pipe and provide frameworks and solutions that are responsive to them,” Illes said.