Novel connectivity map normalization procedure for improved quantitative investigation of structural thalamic connectivity in temporal lobe epilepsy patients
Journal of Magnetic Resonance Imaging
URL with Digital Object Identifier
Background Connectivity studies targeting the thalamus have revealed patterns of atrophy and deafferentiation in temporal lobe epilepsy (TLE). The thalamus can be parcellated using probabilistic tractography to demonstrate regions of cortical connectivity; however, sensitivity to smaller or less connected regions is low. Purpose/Hypothesis To investigate thalamic structural connectivity in a wider range of cortical and limbic structures in TLE patients using a novel connectivity map normalization procedure. Study Type Retrospective. Population/Subjects Patients (N = 23) with medication-resistant TLE and 34 healthy age-matched controls. Field Strength/Sequence For T-1 and diffusion weighting a spoiled gradient sequence was used (41 gradient directions [b = 1000]). For T-2 mapping balanced steady-state free precession was used. Images were acquired at 3T. Assessment Probabilistic tractography and a novel normalization procedure allowed comparison of groups with respect to thalamic connected volume, quantitative MRI, and diffusion tensor imaging (DTI) metrics. Statistical Tests Independent samples t-test, Cohen's d, and Mann-Whitney tests. Results Following normalization, significant differences in thalamic connected volumes were found in left TLE vs. controls bilaterally within the posterior parahippocampal gyrus (L: P = 0.007, confidence interval [CI]: [173.306,1044.41], effect size [ES] = 1.072; R: P = 0.017, CI: [98.677,947.653], ES = 0.945), and contralaterally in the anterior temporal neocortex (P = 0.01, CI: (-2348.09, -333.719), ES = -1.021). This procedure revealed differences in thalamic connected volumes, where previously published procedures could not, and provided a basis for exploratory analysis of quantitative MRI and DTI metrics. Data Conclusion The novel connectivity map normalization scheme proposed here successfully allowed comparison between a wider range of cortical and limbic structures. Multiple volumetric and quantitative MRI and DTI-related differences between TLE patients and controls were revealed following normalization. With validation from a larger cohort, thalamo-temporal connection aberrancies may become useful biomarkers of disease states and probabilistic tractography as a procedure for identification of thalamic targets in modulatory therapies for TLE.