Data mining MR image features of select structures for lateralization of mesial temporal lobe epilepsy

Abstract

PURPOSE: This study systematically investigates the predictive power of volumetric imaging feature sets extracted from select neuroanatomical sites in lateralizing the epileptogenic focus in mesial temporal lobe epilepsy (mTLE) patients. METHODS: A cohort of 68 unilateral mTLE patients who had achieved an Engel class I outcome postsurgically was studied retrospectively. The volumes of multiple brain structures were extracted from preoperative magnetic resonance (MR) images in each. The MR image data set consisted of 54 patients with imaging evidence for hippocampal sclerosis (HS-P) and 14 patients without (HS-N). Data mining techniques (i.e., feature extraction, feature selection, machine learning classifiers) were applied to provide measures of the relative contributions of structures and their correlations with one another. After removing redundant correlated structures, a minimum set of structures was determined as a marker for mTLE lateralization. RESULTS: Using a logistic regression classifier, the volumes of both hippocampus and amygdala showed correct lateralization rates of 94.1%. This reflected about 11.7% improvement in accuracy relative to using hippocampal volume alone. The addition of thalamic volume increased the lateralization rate to 98.5%. This ternary-structural marker provided a 100% and 92.9% mTLE lateralization accuracy, respectively, for the HS-P and HS-N groups. CONCLUSIONS: The proposed tristructural MR imaging biomarker provides greater lateralization accuracy relative to single- and double-structural biomarkers and thus, may play a more effective role in the surgical decision-making process. Also, lateralization of the patients with insignificant atrophy of hippocampus by the proposed method supports the notion of associated structural changes involving the amygdala and thalamus.