Why brain scans need better diversity data to work for everyone

Brain scans like MRIs help doctors spot brain changes linked to diseases. But these scans may not work the same for everyone. Studies show brain measurements can differ widely across ethnic groups and income levels. Yet most brain research is done on a small slice of the population, mainly white and wealthy volunteers. This could lead to bad guesses about what’s normal or abnormal in other groups. Researchers looked at 1, 013 brain imaging studies but only 14 met their strict rules. Those 14 studies showed real differences in brain structure between groups. Some groups had thicker brain layers, others had bigger or smaller deep brain parts. These differences aren’t just random—they’re tied to people’s backgrounds and living conditions.The big problem? Doctors still use one-size-fits-all brain scan rules. If those rules come from one group, they might miss signs of trouble in others. It’s like using a shoe size chart made for adults to fit kids—it just doesn’t work. The review suggests we need new rules made for each population group. AI tools that read brain scans could help, but they’ll be just as biased if trained on limited data. Imagine teaching a computer to recognize dogs using only pictures of poodles—it’ll struggle with bulldogs. Same idea here. Better data from more diverse groups could make these tools fairer and more accurate.