Guessing Brain Age: Why Two Scans Are Better Than One

Ever wished we could tell someone's age just by looking at their brain scan? Scientists often use T1-weighted structural MRI to do this. But a new study mixed things up by combining structural MRI and diffusion MRI. They created a smart model that could see both the big brain features and tiny details. This model was tested on people from two big groups: the Human Connectome Project and the Cambridge Center for Aging and Neuroscience. It was also tested on three other groups to check its accuracy over time. Plus, they looked at how well the model predicted brain age related to real-life skills like walking and feelings of loneliness.

So, what happened? The combined model was way more accurate. It had a mean absolute error of just 2. 44 years for one group and 4. 36 years for the other. That's pretty close! It also did a better job explaining age-related differences in brain scans than single-type models. When tested on other groups, it reduced errors by up to 76%! While the combined model was great, the single-type model was a bit better at predicting changes over time. But here's where it gets interesting: the combined model found links between predicted brain age and behaviors that we might not see otherwise. For instance, people with older-looking brains might have less walking endurance or feel more lonely. By using this clever combination, scientists can get a clearer picture of brain aging and understand how it affects our daily lives.

Actions