Harmonized Threshold Free Network-based Statistics for Characterization of Neuropsychological and White Matter Connectome Relationships in Patients with Temporal Lobe Epilepsy
Abstract number :
2.171
Submission category :
5. Neuro Imaging / 5A. Structural Imaging
Year :
2022
Submission ID :
2204596
Source :
www.aesnet.org
Presentation date :
12/4/2022 12:00:00 PM
Published date :
Nov 22, 2022, 05:25 AM
Authors :
Daniel Chu, BA – University of Wisconsin School of Medicine and Public Health; Nagesh Adluru, PhD – University of Wisconsin Madison; Veena Nair, PhD – Radiology – University of Wisconsin Madison; Anusha Adluru, MS – University of Wisconsin Madison; Kevin Dabbs, MS – University of Wisconsin Madison; Jedidiah Mathis, BS – Medical College of Wisconsin; Andrew Nencka, PhD – Medical College of Wisconsin; bruce Hermann, PhD – University of Wisconsin Madison; Andrew Alexander, PhD – University of Wisconsin Madison; Aaron Struck, MD – University of Wisconsin Madison; Jeffrey Binder, MD – Medical College of Wisconsin; Mary Meyerand, PhD – University of Wisconsin Madison; Vivek Prabhakaran, MD PhD – University of Wisconsin Madison
Rationale: Our study aims to characterize the aberrant white-matter connectivity in temporal lobe epilepsy (TLE) patients compared to healthy controls (HCs) in a homogenous large-scale patient study from the Epilepsy Connectome Project (ECP). We elucidated how generalized tonic-clonic (GTC) seizures differentially affect specific connections in the white-matter connectome. Furthermore, we investigated whether these connections are differentially related to the cognitive outcomes measured using auditory verbal learning test (AVLT), as well as cluster-based analysis of cognitive impairment status.
Methods: Multi-shell connectome diffusion weighted MRI (ms-dMRI) data from 174 subjects ages 18 to 60 were employed, including 85 TLE patients (34 male, mean age = 39.28 ± 11.71 years) and 39 HCs (20 male, mean age = 34.87 ± 10.20 years. The ms-dMRI was pre-processed using the DESIGNER guidelines. The IIT Destrieux gray matter atlas was used to derive the 162 x 162 structural connectivity matrices (SCMs) using MRTrix3. ComBat data harmonization was applied to the structural connectivity matrices (SCMs) from pre- and post-scanner upgrade acquisitions and threshold free network-based statistics (TFNBS) was used for statistical analysis of the harmonized SCMs. General linear model (GLM) was used to test the difference in white matter connectomes between the GTC and TLE patients who never had GTC. GLMs were also used to test the white-matter connections that may be differentially related to cognitive profiles including three clusters of intact cognition, focal cognitive impairment, and generalized cognitive impairment within the TLE group. Additionally, the effect of interaction between AVLT (total standardized score and delayed recall score), a verbal memory assessment to evaluate the severity of memory dysfunction, and the epilepsy status (TLE vs. HC) was tested. For all the GLMs, age and sex were used as confounding variables.
Results: Our results demonstrate that age, sex, and epilepsy onset age adjusted mean cross-sectional area (CSA) is consistently lower across the representative 16 significant tracts in TLE patients who experienced GTCs compared to TLE patients who never had a GTC. This illustrates that the occurrence of a GTC may have specifically impacted these white-matter tracts (Figure 1). In addition, our summary neuropsychological testing is depicted in Figure 2.
Conclusions: Overall, these results reveal that presence of GTC seizures can cause abnormalities of the white matter connectivity abnormalities in 16 tracts, as well as discernable differences in graph theoretic measures of connectivity and network-based statistics. In addition, correlations with neuropsychological assessments and cognitive impairment clusters further provide elucidation of white matter tract abnormalities in TLE. Ultimately, these characterizations contribute to the identification of novel MRI biomarkers, further substantiating that TLE is a progressive network disorder involving ongoing neuroplasticity.
Funding: We are grateful for the support from the AES Pre-doctoral Fellowship, MSTP Grant: T32 GM140935, UW MSTP Radiology Fellowship, and NIH grants R01NS123378, R01NS105646, R01NS105646, R01NS111022, and P50HD105353.
Neuro Imaging