Using Zeffiro Interface to optimize current density in Deep Brain Stimulation
Lassila, Antti (2023)
Lassila, Antti
2023
Teknis-luonnontieteellinen DI-ohjelma - Master's Programme in Science and Engineering
Tekniikan ja luonnontieteiden tiedekunta - Faculty of Engineering and Natural Sciences
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2023-08-03
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-202306266906
https://urn.fi/URN:NBN:fi:tuni-202306266906
Tiivistelmä
This thesis aims at implementing a mathematical optimization exercise for Deep Brain Stimulation (DBS) for determining optimal current injection among the neurostimulator's electrodes using backpropagation and the recently implemented L1-norm regularized L1-norm fitting (L1L1) optimization method. The behaviour of the elicited current density and field ratio.
The goal is to develop a framework where the user can generate a volumetric mesh of a realistic head model and investigate DBS lead designs and current directionality, allowing one to observe the mutual difference between different electrode leads and how the resolution of the points of interest affects the solution.
The thesis presents the mathematical background of deep brain stimulation and how to create a lead field matrix using Maxwell's equations as well as the optimization algorithms. The main windows in Zeffiro Interface are presented, as to what needs to be done in Zeffiro Interface to create the electrode lead, and then how the optimal solutions were calculated.
The open Matlab-based Zeffiro Interface (ZI) toolbox implemented all activities and scripts. The obtained solutions showed that a bipolar version (one anodal, one cathodal) did not provide a high field ratio compared to a multi-channel DBS lead, such as Medtronics-sapiens or Abbott Infinity, both with 40 and 8 contact points, respectively. Increasing the points of interest made the solutions more stable concerning the nuisance weight minimum values and the focused current density was higher in most of the cases. Still, the calculation time increased significantly as the size of the lead field matrix grew.
The goal is to develop a framework where the user can generate a volumetric mesh of a realistic head model and investigate DBS lead designs and current directionality, allowing one to observe the mutual difference between different electrode leads and how the resolution of the points of interest affects the solution.
The thesis presents the mathematical background of deep brain stimulation and how to create a lead field matrix using Maxwell's equations as well as the optimization algorithms. The main windows in Zeffiro Interface are presented, as to what needs to be done in Zeffiro Interface to create the electrode lead, and then how the optimal solutions were calculated.
The open Matlab-based Zeffiro Interface (ZI) toolbox implemented all activities and scripts. The obtained solutions showed that a bipolar version (one anodal, one cathodal) did not provide a high field ratio compared to a multi-channel DBS lead, such as Medtronics-sapiens or Abbott Infinity, both with 40 and 8 contact points, respectively. Increasing the points of interest made the solutions more stable concerning the nuisance weight minimum values and the focused current density was higher in most of the cases. Still, the calculation time increased significantly as the size of the lead field matrix grew.