TY - GEN
T1 - Automatic identification of DBS parameters from the volume of tissue activated (VTA) using support vector machines
AU - Aguilar, Robinson
AU - Vargas-Cardona, Hernán Darío
AU - Álvarez, Andrés M.
AU - Orozco, Álvaro A.
AU - Navarro, Piedad
N1 - Publisher Copyright:
© Springer Nature Switzerland AG 2019.
PY - 2019
Y1 - 2019
N2 - Deep brain stimulation (DBS) is a neurosurgical method to treat symptoms of Parkinson’ disease. Several computational models, mostly based on finite element method (FEM) have been employed to describe the interaction of electromagnetic waves in brain tissues during DBS. Also, for planning the DBS, it is necessary to estimate with precision the neural response generated by electrodes in the stimulated region, what it is known as volume of tissue activated (VTA). However, this estimation should consider the intrinsic properties of each patient, therefore DBS parameters must be adjusted individually. In this work, we propose a 3D interaction module for estimating the DBS parameters (amplitude, contacts, among others) from a desired VTA using support vector machines (inverse problem). Also, we developed an interactive application for analyzing the VTA generated by DBS in the subthalamic nucleus (STN) combining medical imaging and non-rigid deformation models. This module is a part of the NEURONAV software, previously developed for clinical support during postoperative therapy of neuro-modulation performed in Colombian PD patients. Outcomes show that it is possible to estimate with high accuracy the DBS parameters for different subjects.
AB - Deep brain stimulation (DBS) is a neurosurgical method to treat symptoms of Parkinson’ disease. Several computational models, mostly based on finite element method (FEM) have been employed to describe the interaction of electromagnetic waves in brain tissues during DBS. Also, for planning the DBS, it is necessary to estimate with precision the neural response generated by electrodes in the stimulated region, what it is known as volume of tissue activated (VTA). However, this estimation should consider the intrinsic properties of each patient, therefore DBS parameters must be adjusted individually. In this work, we propose a 3D interaction module for estimating the DBS parameters (amplitude, contacts, among others) from a desired VTA using support vector machines (inverse problem). Also, we developed an interactive application for analyzing the VTA generated by DBS in the subthalamic nucleus (STN) combining medical imaging and non-rigid deformation models. This module is a part of the NEURONAV software, previously developed for clinical support during postoperative therapy of neuro-modulation performed in Colombian PD patients. Outcomes show that it is possible to estimate with high accuracy the DBS parameters for different subjects.
UR - http://www.scopus.com/inward/record.url?scp=85063043250&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-13469-3_86
DO - 10.1007/978-3-030-13469-3_86
M3 - Conference contribution
AN - SCOPUS:85063043250
SN - 9783030134686
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 740
EP - 747
BT - Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications - 23rd Iberoamerican Congress, CIARP 2018, Proceedings
A2 - Vera-Rodriguez, Ruben
A2 - Fierrez, Julian
A2 - Morales, Aythami
PB - Springer Verlag
T2 - 23rd Iberoamerican Congress on Pattern Recognition, CIARP 2018
Y2 - 19 November 2018 through 22 November 2018
ER -