Most importantly, this is just the beginning! To estimate DunedinPACNI in your data, all you need is a single T1-weighted structural MRI and our publicly available tool. Please use it and share it widely! github.com/etw11/Dunedi...

Across several datasets, faster DunedinPACNI was associated with poor cognition, physical frailty, poor health, and worse cognitive status. Furthermore, faster DunedinPACNI predicted faster hippocampal atrophy, earlier onset of chronic disease, dementia, and mortality.

Building on the epigenetic clock and brain-age literatures, we built a "next-generation" brain aging clock by predicting an individual's rate of longitudinal biological aging from a single brain scan.

We call our measure "DunedinPACNI"

Here's the link - doi.org/10.1038/s435...

This work was an amazing collaboration with my co-first author @ethanwhitman777.bsky.social, as well as a great team at Duke and Otago Annchen Knodt, Av Caspi, Terrie Moffitt, and Ahmad Hariri

This was a massive team effort - @jingnandu.bsky.social, @jaredniels.bsky.social, Randy Buckner and many others!

The high precision afforded by cluster scanning promises to accelerate clinical trials, lead to personalized biomarkers, and be a useful tool for the science of individual differences in brain development, aging, and disease. Check out the preprints for more details!

This precision allowed us to see clear deviations from expected aging when they arose. In a striking example, we detected an aggressive atrophy trajectory in an individual who was cognitively unimpaired at baseline but then went on to develop an MCI diagnosis.

Critically, this reduction in error allowed us to see changes in individuals within just one year. Here are hippocampal aging trajectories in 8 individuals across one year where you can see large individual differences and the benefits of pooling multiple measurements

We found a solution - cluster scanning. Utilizing the latest advances in scan acceleration we collected several short, 1-minute long T1s to drive to measurement error within individuals ... and it worked!

A key challenge for brain aging is to measure brain changes in individuals rather than group averages. This has huge implications for clinical trials, basic aging research and individual differences. However, we lack the precision needed to detect changes over short intervals.