Quantitative volume-based morphometry in focal cortical dysplasia: A pilot study for lesion localization at the individual level

Purpose: Surgical resection is the most effective treatment for focal cortical dysplasia (FCD). However, many patients with FCD have unremarkable or even negative findings on conventional magnetic resonance imaging (MRI). In this study, we explored the brain volume abnormalities of FCD patients at the individual level using an experimental volume-based morphometry algorithm and further estimated whether the volume abnormalities can help in the detection of FCD lesions. Materials and methods: Sixteen patients with histologically-proven FCD lesions were retrospectively studied. Among them, eight patients had no visible abnormalities on routine MRI, three had abnormalities which partly matched the location of the surgical resection regions, and two did not match. For each patient, cerebral high-resolution Tl-weighted magnetization-prepared rapid acquisition with gradient echo (MPRAGE) images were segmented into 45 structures, according to a brain anatomy template, and the volume of each structure was compared with an age- and gender-matched normal population at the individual level, based on a MorphoBox prototype. A Receiver Operating Characteristics (ROC) curve was used to evaluate the performance of the prototype in patients. To find the most appropriate threshold value for localizing the epileptogenic zones, deviations from the normative ranges of each resulting volume estimate were assessed by z-scores. Results: Volume abnormalities including atrophic and hypertrophic volumes could be found in all the patients. Epileptogenic zones were found in brain structures with an abnormal volume in 87.5% (14/16) of patients. In 71.4% of patients (10/14), these zones were fully located in regions with an atrophic volume. This suggests that FCD lesions are more likely to be in regions with an atrophic volume than in those with a hypertrophic volume. When the best cut-off z-score value was -3.0, the sensitivity, specificity, and ROC area under the curve of the volume estimates were 93.9%, 79.6%, and 0.89, respectively. Conclusion: Volume abnormalities can assist in the diagnosis of epileptogenic zones at the individual level in FCD patients with negative or positive findings on conventional MR images. Atrophic regions are more likely than hypertrophic ones to represent epileptogenic zones. Volume-based morphometry based on a MorphoBox prototype has potential to assist a careful scrutiny by radiologists with target in atrophic regions in patients who are initially deemed to be MR-negative, further trying to increase the detection rate of FCD.