Apostolos Mikroulis presents Patient-specific and interpretable deep brain stimulation optimisation using MRI and clinical review
On 2025-11-13 11:00:00 at G205, Karlovo náměstí 13, Praha 2
Optimisation of Deep Brain Stimulation (DBS) settings is essential for
achieving
clinical efficacy in movement disorders such as Parkinson’s disease. While
modern approaches often rely on data-intensive statistical or machine learning
techniques that add significant overhead to clinical workflows, we present a
geometry-based optimisation method for DBS electrode contact and current
selection, grounded in routinely collected MRI data, established software tools
(Lead-DBS), and optionally, clinical review records. The proposed pipeline,
implemented as a cross-platform tool, utilises lead reconstruction data and
simulations of the Volume of Tissue Activated (VTA) to estimate contacts
optimally positioned relative to the target structure and to suggest
appropriate
stimulation currents, allowing for interactive fine-tuning and optional
integration of existing clinical evaluations to avoid adverse effects. Using
174
electrode reconstructions from 87 Parkinson’s disease patients, our results
show that the algorithm’s DBS parameter settings achieve significantly better
target coverage (Wilcoxon p < 5e−13, Hedges’ g > 0.94) and reduced electric
field spread to neighbouring regions (p < 2e−10, g > 0.46) compared with
expert-defined settings. Retrospective analysis of a subset of 50 cases further
predicts motor outcomes comparable to expert settings (g = 0.05–0.08, p =
0.09–1), indicating potential for similar clinical efficacy pending
prospective validation. In conclusion, this automated, geometry-based method
demonstrates strong performance and seamless integration into current clinical
workflows, providing an efficient alternative to manual and machine
learning–based DBS parameter optimisation [1].
[1]Mikroulis Apostolos , Lasica Andrej , Filip Pavel , Bakstein Eduard , Novak
Daniel, Patient-specific and interpretable deep brain stimulation optimisation
using MRI and clinical review, Frontiers in Neuroscience, Volume 19 - 2025,
https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2025.1661987
achieving
clinical efficacy in movement disorders such as Parkinson’s disease. While
modern approaches often rely on data-intensive statistical or machine learning
techniques that add significant overhead to clinical workflows, we present a
geometry-based optimisation method for DBS electrode contact and current
selection, grounded in routinely collected MRI data, established software tools
(Lead-DBS), and optionally, clinical review records. The proposed pipeline,
implemented as a cross-platform tool, utilises lead reconstruction data and
simulations of the Volume of Tissue Activated (VTA) to estimate contacts
optimally positioned relative to the target structure and to suggest
appropriate
stimulation currents, allowing for interactive fine-tuning and optional
integration of existing clinical evaluations to avoid adverse effects. Using
174
electrode reconstructions from 87 Parkinson’s disease patients, our results
show that the algorithm’s DBS parameter settings achieve significantly better
target coverage (Wilcoxon p < 5e−13, Hedges’ g > 0.94) and reduced electric
field spread to neighbouring regions (p < 2e−10, g > 0.46) compared with
expert-defined settings. Retrospective analysis of a subset of 50 cases further
predicts motor outcomes comparable to expert settings (g = 0.05–0.08, p =
0.09–1), indicating potential for similar clinical efficacy pending
prospective validation. In conclusion, this automated, geometry-based method
demonstrates strong performance and seamless integration into current clinical
workflows, providing an efficient alternative to manual and machine
learning–based DBS parameter optimisation [1].
[1]Mikroulis Apostolos , Lasica Andrej , Filip Pavel , Bakstein Eduard , Novak
Daniel, Patient-specific and interpretable deep brain stimulation optimisation
using MRI and clinical review, Frontiers in Neuroscience, Volume 19 - 2025,
https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2025.1661987