Clinical Study Voxel-based morphometry in the detection of dysplasia and neoplasia in childhood epilepsy: Limitations of grey matter analysis Jason M. Bruggemann a , Marko Wilke b , Seu S. Som a , Ann M.E. Bye a , Andrew Bleasel c , John A. Lawson a, * a Department of Neurology, Sydney Children’s Hospital and University of New South Wales, High Street, Randwick, New South Wales 2031, Australia b Department of Pediatric Neurology and Developmental Medicine, Children’s Hospital, University of Tuebingen, Germany c Department of Neurophysiology, The Children’s Hospital at Westmead, Westmead, New South Wales, Australia article info Article history: Received 25 October 2007 Accepted 21 August 2008 Keywords: Epilepsy Children Dysplasia Neoplasia Voxel-based morphometry Statistical parametric mapping abstract The purpose of this exploratory investigation was to evaluate voxel-based morphometry (VBM) in detect- ing lesions underlying childhood epilepsy, and to establish the optimal image processing and statistical parameters in this context. The patients were 16 children (10 boys) aged 5.9 to 15.2 years (mean 11.3 years) with epilepsy and focal cortical dysplasia (FCD) or neoplasia. The control group comprised 24 nor- mal children (12 boys), age matched to the patients. MRI volumes were spatially normalised to a custom template and segmented into grey matter (GM) and white matter. Using statistical parametric mapping, the GM segment from each patient was then contrasted with the mean GM segment of the control group utilising different VBM post-processing methods. Maps showing increased/decreased areas of GM con- centration or volume were generated and compared with visually identified lesions. The results indicated that conservative VBM parameters of linear normalisation with no modulation produced the highest rates of lesion detection, which were identical for FCD and neoplasia at 5/8 lesions. These preliminary data suggest that VBM analysis of GM using conservative parameters can usually detect FCD and neopla- sia in the MRI of children with epilepsy, but sensitivity may be inadequate for routine clinical application. Further refinement of the technique may be necessary. Crown Copyright Ó 2008 Published by Elsevier Ltd. All rights reserved. 1. Introduction Intractable childhood epilepsies are often caused by cortical dysplasia and neoplasia. 1–6 Visual interpretation of MRI can be effectual in identifying focal cortical dysplasia (FCD), which is evi- dent on MRI as thickening and signal change of the cortex, abnor- mal gyral patterns, blurring of the grey matter (GM)–white matter (WM) margins, and WM signal changes. 5–9 However, visual identi- fication of FCD on MRI also has limitations because subtle manifes- tations may remain undetected and the histological abnormality may be greater than the region apparent on MRI. 1,4,6–13 Further, vi- sual interpretation of MRI is highly subjective and reflects the skill of the radiologist. 14–16 Thus, there is a compelling need to explore objective methods that may assist in the precise identification of FCD on MRI. Voxel-based morphometry (VBM) may have a role in clinical assessment. 12–15,17–24 VBM is an automated and objective, voxel- by-voxel examination of regional brain tissue concentration and volume in high-resolution structural MRI. 25–28 VBM has been em- ployed in the detection and delineation of lesions in individual epilepsy patient versus control-group comparisons. 12–15,20–24,29 In terms of malformations of cortical development (MCD; including FCD), VBM can detect increases in GM concentration (GMC) con- cordant with visible lesions on MRI and other seizure localisation data. 12–15,23 There is also some evidence that it can identify more subtle GM lesions in examples of reportedly normal MRI. 18,21,22,30 Nonetheless, the value of VBM as a clinical tool is yet to be defin- itively ascertained; it may lack the sensitivity and specificity re- quired for accurately determining lesions. 31 VBM has been utilised for characterising differences in GM and WM between groups of paediatric epilepsy patients and controls, 32,33 yet there is little published data concerning the potential clinical use of VBM with MRI from individual paediatric epilepsy patients. This is an exploratory investigation to assess the viability of automated VBM analysis of structural MRI for detecting lesions underlying childhood epilepsy. Although our primary interest was children with FCD and epilepsy, we also included a group of children with epilepsy and tumours for comparison purposes be- cause neoplasia is generally easier to identify in MRI than dyspla- sia. 34 The specific aim was to determine the efficacy of VBM for identifying abnormalities of GMC or GM volume (GMV) and to establish the optimal processing and statistical parameters in this context. A more limited investigation was also conducted with WM segments in regard to WM concentration (WMC). 0967-5868/$ - see front matter Crown Copyright Ó 2008 Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jocn.2008.08.025 * Corresponding author. Tel.: +61 2 9382 1658; fax: +61 2 9382 1580. E-mail address: john.lawson@unsw.edu.au (J.A. Lawson). Journal of Clinical Neuroscience 16 (2009) 780–785 Contents lists available at ScienceDirect Journal of Clinical Neuroscience journal homepage: www.elsevier.com/locate/jocn