Abstract
It is crucial to differentiate patients with temporal lobe epilepsy (TLE) from the healthy population and determine abnormal brain regions in TLE. The cortical features and changes can reveal the unique anatomical patterns of brain regions from structural magnetic resonance (MR) images. In this study, structural MR images from 41 patients with left TLE, 34 patients with right TLE, and 58 normal controls (NC) were acquired, and four kinds of cortical measures, namely cortical thickness, cortical surface area, gray matter volume (GMV), and mean curvature, were explored for discriminative analysis. Three feature selection methods including the independent sample t-test filtering, the sparse-constrained dimensionality reduction model (SCDRM), and the support vector machine-recursive feature elimination (SVM-RFE) were investigated to extract dominant features among the compared groups for classification using the support vector machine (SVM) classifier. The results showed that the SVM-RFE achieved the highest performance (most classifications with more than 84% accuracy), followed by the SCDRM, and the t-test. Especially, the surface area and GMV exhibited prominent discriminative ability, and the performance of the SVM was improved significantly when the four cortical measures were combined. Additionally, the dominant regions with higher classification weights were mainly located in the temporal and the frontal lobe, including the entorhinal cortex, rostral middle frontal, parahippocampal cortex, superior frontal, insula, and cuneus. This study concluded that the cortical features provided effective information for the recognition of abnormal anatomical patterns and the proposed methods had the potential to improve the clinical diagnosis of TLE.
Highlights
Epilepsy, affecting approximately 50 million patients worldwide, has attracted increasing attention of many investigators
10 of 41 left temporal lobe epilepsy (LTLE) and 8 of 34 right TLE (RTLE) patients were not identified in magnetic resonance (MR) scans
Three feature selection methods were investigated for detection of temporal lobe epilepsy (TLE) using cortical surface features from brain structural magnetic resonance imaging (MRI)
Summary
Epilepsy, affecting approximately 50 million patients worldwide, has attracted increasing attention of many investigators. There were differences in the extent of anatomical damage between hemispheres [5, 6] and the morphological abnormalities were more widespread in the left temporal lobe epilepsy (LTLE) with gray matter volume (GMV) loss, especially in the hippocampus, the parahippocampal gyrus, and the entorhinal cortex [7]. Morphological features of the cerebral cortex such as cortical thickness (CTh), surface area, GMV, and mean curvature (MCu) have been found to be associated with pathogenesis of either the LTLE or the RTLE. Previous studies indicated that there existed asymmetric reduction of cortical surface area (CSA) in the ipsilateral mesial and the anterior temporal lobe subregions [16], MCu abnormality in the bilateral insula [17], overall cortical thinning [18], and gray volume loss [19]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
