Automated detection of gray matter malformations using optimized voxel-based morphometry: a systematic approach

Abstract

Malformations of cortical development (MCD) are a recognized cause of epilepsy. Their special significance lies in the fact that, once detected and delineated, they are amenable to surgical removal. However, diagnosis from high-resolution MRI is still difficult, time-consuming, and highly dependent on individual expertise. We have recently proposed a simple procedure to detect cortical dysplasias, using automated procedures available within SPM99 (Wellcome Department, University College London, UK). Here, we aimed to systematically determine the best combination of processing parameters, using an optimized voxel-based morphometry approach. We included 20 patients with a known MCD and compared them to a normal database of 53 healthy, age- and gender-matched controls. The approaches taken during spatial normalization and a number of other parameters were systematically altered in order to find the best combination of parameters. Overall, 99 different approaches were evaluated in different ways. As far as possible, automatic processing and evaluation steps were used. With the number of candidate regions for MCD limited to five per patient, the best approaches resulted in the correct identification of up to 16 of 20 malformations. However, a number of approaches failed to perform well. The reasons for these failures and the implications this has for other studies are discussed. We conclude that voxel-based morphometry is able to detect cortical malformations with a high degree of accuracy. However, specific problems seem to arise when using an optimized protocol for voxel-based morphometry, indicating that this protocol may not be optimal for all voxel-based studies on brain morphology. Our approach, involving systematic alterations of parameters and evaluation, may be useful for other studies.

Introduction

Malformation of cortical development (MCD) are a recognized cause of chronic epilepsy. Their special significance lies in the fact that they are amenable to treatment by surgical resection and that they are increasingly diagnosed due to the more common use of high-resolution magnetic resonance imaging (MRI) Andermann, 2000, Barkovich et al., 2001, Sisodiya, 2000. However, diagnosis is time-consuming and difficult, and standard MRI can be unrevealing in a high percentage of patients (Tassi et al., 2002). Especially in subtle cases, diagnosis relies heavily on the experience of the radiologist reading the images (Hong et al., 2000). In order to address this problem, a number of attempts have been made to employ modern postprocessing image strategies to aid in the diagnosis of these disorders Bastos et al., 1999, Woermann et al., 1999.

We have recently suggested a simple, voxel-based technique to detect cortical dysplasias by comparison with a healthy control group (Kassubek et al., 2002). We aimed at systematically optimizing this approach in a large patient and control population by (1) applying a newly suggested, optimized processing stream for structural imaging data (Good et al., 2001) and testing its suitability and (2) finding the combination of processing parameters that offers the highest rate of detection of MCD and the best distinction between MCD patients and healthy controls. In order to avoid possible user errors and to minimize the amount of time necessary, all procedures and evaluations should be automated as far as possible.