Matthieu Bernard

11/
If you're interested in aging, navigation, attractor networks, and computational neuroscience, we’d love your feedback!
Thanks to Jonathan Shine, @wolberslab.bsky.social & @andrejbicanski.bsky.social for the supervision and collaboration.
Preprint: www.biorxiv.org/content/10.1...

10/
Why this matters:
The HD system may be an early marker of navigational decline and potentially of neurodegenerative change.
Understanding how aging affects HD attractor dynamics opens new avenues for translational work—from rodent aging studies to clinical interventions.

9/
This framework offers:
- Direct human evidence that the HD system is impacted by aging
- A mechanistic account of directional drift
- Support for attractor models that withstand perturbations
- Implications for memory, decision-making, and other attractor-based functions.

8/
The key insight:
Synaptic changes + vestibular noise + modest neuron loss best reproduce the behavioral drift we see in older adults.
These factors create shallow attractor basins → the HD bump becomes less stable → orientation estimates drift, especially with time delays.

7/
We then introduced aging-like perturbations:
- Synaptic degradation
- Increased noise in vestibular signals
- Neuron deletion
Each produces distinct signatures in HD stability, letting us test mechanistic hypotheses.

6/
Behavior alone can’t reveal why this drift occurs.
So we built a biologically realistic attractor-network model of the HD system that removes direct recurrence, uses persistent vestibular drive, and stays stable under perturbations—a long-standing challenge in attractor theory.

5/
Key findings: Performance drops when delay is added, but older adults show a much larger decline with systematic drift in orientation estimates. Errors build over time.
This is the first direct evidence that HD computations degrade with age in humans.

4/
We built an immersive VR orientation task that isolates HD processing.
Participants view a distant landmark, rotate, and must reproduce their facing direction — sometimes after a 20-second delay.
This setup allows us to probe the stability of internal directional estimates.

3/
That system is the head direction (HD) network — the brain’s internal compass.
Rodent studies suggest aging alters HD representations, but direct human evidence was missing.
So we set out to test HD computations in older vs. younger adults.

2/
Spatial navigation is one of the first cognitive abilities to decline with age—and often early in conditions like dementia.
But the neural mechanisms behind this decline remain unclear.
Most work has focused on the medial temporal lobe… but one key system has been overlooked.

Excited to share my first PhD preprint!
“How human aging disrupts the head direction network: evidence from VR experiments and mechanistic models”
We investigated why our sense of direction becomes less stable as we age.
Preprint: www.biorxiv.org/content/10.1...
1/ thread 🧵