English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Adaptive deep brain stimulation (aDBS) can profit from data-driven approaches developed for BCIs. These aDBS systems improve upon the constant DBS in terms of efficiency and side effects amelioration by taking the ongoing brain state into consideration. The environment controlled by aDBS is governed by partial observability, rendering classic control strategies suboptimal. In this regard, development of novel approaches is critical for improved aDBS therapy. However, early stage aDBS development is a difficult endeavor, given the lack of suitable development platforms. In our contribution, we present a simulated environment that allows to modularly embed different surrogates of key challenges found in the aDBS problem. Specifically, we will focus on partial observability stemming from non-stationary dynamics and noisy state representations. Our simulations are used to analyze representative reinforcement learning approaches regarding their ability to cope with the partial observability. To allow reproducibility and encourage adoption of our approach, the source code of our experiments is made available online.

A Simulated Environment for Studying Partial observability in Novel adaptive Deep brain stimulation.