Modeling between DBS signals and brain white matter integrity based Dystonia patients
Poster No:
27
Submission Type:
Abstract Submission
Authors:
Sumiko Abe1, Maral Kasiri2, Alireza Mousavi2, Estefania Hernandez-Martin3, Terence Sanger2
Institutions:
1CHOC Children's Hospital, City of Orange, CA, 2University of California, Irvine, Irvine, CA, 3Universidad de La Laguna, La Laguna, Santa Cruz de Tenerife
First Author:
Co-Author(s):
Introduction:
Dystonia is a disorder of motor programmes controlling semiautomatic movements or postures, with clinical features such as sensory trick, which suggests sensorimotor mismatch as the basis. Dystonia was originally classified as a basal ganglia disease. it is now regarded as a 'network' disorder[1]. Dystonia is a form of dyskinetic cerebral palsy(CP), and CP is associated with white matter injury[3]. We propose that DBS signals travel through white matter tracts to affect both local and distant brain sites. Deep brain stimulation (DBS) has been an important treatment for movement disorders, such as dystonia or Parkinson's disease[2-3]. we hypothesize that the signal of DBS transfer is related with the white matter injury such as demyelination, axonal degeneration etc. DTI is one of the most important method for quantifying these damages of white matter. In this study, we built a general linear modelGML) to descripe the relationships between the DBS signal and DTI coefficients. The result shows the DBS signal peak-to-peak amplitude (P2P) and time-to-(first-) peak delay (T2P) are related with fiber length, fiber diameter and fractional Anisotropy.
Methods:
Neuroimages of 6 children with dystonia are used in this study. Our clinical procedure for determining DBS targets includes the implantation of 10 temporary AdTech MM16C depth electrodes (Adtech Medical Instrument Corp., Oak Creek, WI, USA) at potential DBS targets (including basal ganglia and thalamic subnuclei), as identified based on clinical criteria in each patient. Electrophysiology recordings were performed during the first 24 to 48 hours after clinical implantation of the temporary stereo-electroencephalography (sEEG) depth electrodes.
Diffusion images were processed using the TOP-UP for motion artifacts were corrected through Eddy-current corrections. After correcting the distortions, both post-surgery CT and DWI volumes were aligned to the same plane using the FLIRT tool in FSL. The voxel size of the T1-weighted images was re-sliced to 1mm3 and used as the source image to warp the CT and DTI images. As a result, the pre-surgery DWI and post-surgery CT volumes were aligned to the structuralMRI volume (T1-weighted) for each subject.
The DTI coefficients, such as tract length, tract diameter, and fractional anisotropy (FA), can be used to quantify the characteristics of each fiber tract. On the other hand, the evoked potentials (EPs) can be characterized by their peak-to-peak amplitude (P2P) and time-to-(first-) peak delay (T2P). In this study, our goal was to explore the relationship between these EP characteristics and DTI coefficients. We hypothesize that the fiber length will be correlated with the delay (T2P) while fiber diameter will be inversely correlated with delay and positively correlated with the amplitude (P2P).In order to test our hypotheses at group level we used generalized linear model (GLM). The linear model was defined as:
T 2P = a · F A + b · L + c · R + d · D+ p1
P 2P = e · F A + f · L + g · R + h · D + p2
Diffusion images were processed using the TOP-UP for motion artifacts were corrected through Eddy-current corrections. After correcting the distortions, both post-surgery CT and DWI volumes were aligned to the same plane using the FLIRT tool in FSL. The voxel size of the T1-weighted images was re-sliced to 1mm3 and used as the source image to warp the CT and DTI images. As a result, the pre-surgery DWI and post-surgery CT volumes were aligned to the structuralMRI volume (T1-weighted) for each subject.
The DTI coefficients, such as tract length, tract diameter, and fractional anisotropy (FA), can be used to quantify the characteristics of each fiber tract. On the other hand, the evoked potentials (EPs) can be characterized by their peak-to-peak amplitude (P2P) and time-to-(first-) peak delay (T2P). In this study, our goal was to explore the relationship between these EP characteristics and DTI coefficients. We hypothesize that the fiber length will be correlated with the delay (T2P) while fiber diameter will be inversely correlated with delay and positively correlated with the amplitude (P2P).In order to test our hypotheses at group level we used generalized linear model (GLM). The linear model was defined as:
T 2P = a · F A + b · L + c · R + d · D+ p1
P 2P = e · F A + f · L + g · R + h · D + p2
Results:
As expected, the multimodal analysis shows a significant correlation between the DTI coefficients and electrophysiological characteristics. The EP amplitude (P2P) is positively correlated with FA and tract volume, while EP delay (T2P) exhibits a negative correlation with both measures. Interestingly, EP amplitude shows a negative correlation with tract length, while EP delay do not show a significant relation to tract length.
Conclusions:
Our study has successfully demonstrated the quantification of the relationship between 22-2 electrophysiological signals and white matter integrity using techniques such as DTI and sEEG recordings during DBS. By comparing tractography with ground truth electrophysiology, we have identified significant correlations between neural tract characteristics and neural responses to electrical stimulation in deep brain structures.
Brain Stimulation:
Deep Brain Stimulation 1
Modeling and Analysis Methods:
Connectivity (eg. functional, effective, structural)
Diffusion MRI Modeling and Analysis 2
Image Registration and Computational Anatomy
Keywords:
ELECTROCORTICOGRAPHY
Modeling
MRI
Sub-Cortical
1|2Indicates the priority used for review
Provide references using author date format
[2] Sanger TD, Delgado MR, Gaebler-Spira D, Hallett M, Mink JW; Task Force on Childhood Motor Disorders Classification and definition of disorders causing hypertonia in childhood. Pediatrics 2003; 111(1): e89-e97.
[3] T. D. Sanger and S. N. Kukke, “Abnormalities of tactile sensory function in children with dystonic and diplegic cerebral palsy,” J. Child Neurol., vol. 22, no. 3, pp. 289–293, 2007.