# Real-time multisite invasive neural recording during downhill skiing in Parkinson’s disease: a case report

## TL;DR

This case study shows how neural activity in Parkinson’s disease differs during downhill skiing compared to in-clinic tasks, revealing new patterns linked to movement and medication.

## Contribution

The study presents the first real-time neural recordings during complex outdoor sports in a Parkinson’s patient, uncovering unique frequency patterns.

## Key findings

- Skiing induced STN beta desynchronization and low-gamma activity, distinct from in-clinic tasks.
- Unique 85 Hz finely-tuned gamma activity was observed in the less affected hemisphere during skiing.
- Tremor-related beta suppression and 10 Hz activity were associated with stopping and resting tremor.

## Abstract

Invasive recording of neural activity provides valuable insights into Parkinson’s disease (PD). Bidirectional sensing devices enable wireless neural data collection during everyday activities, but neural signals during complex outdoor sports remain unexplored.

We recorded neural data from a 57-year-old PD patient using bilateral implanted pulse generators connected to subthalamic nucleus (STN) and motor cortex leads. Recordings were performed in two settings: in-clinic during a computer-controlled task and outdoors during downhill skiing. Neural data were analyzed for power spectral density (PSD) and coherence across different frequencies.

In-clinic recordings demonstrated movement-related cortical and STN beta desynchronization with cortical gamma increase. Skiing similarly induced STN beta desynchronization but also elicited low-gamma activity (30–60 Hz) and unique finely-tuned gamma (FTG) activity at 85 Hz in the off-medication state, predominantly in the less affected hemisphere. Tremor-related cortical beta suppression was observed during stopping, with prominent 10 Hz activity associated with resting tremor.

Real-time multisite neural recordings during a complex outdoor activity revealed distinct neural signatures compared to in-clinic tasks. The findings suggest that self-cued, learned motor tasks elicit unique frequency bands and highlight differences based on disease asymmetry and medication state.