Clinical Value of Machine Learning in the Automated Detection of Focal Cortical Dysplasia Using Quantitative Multimodal Surface-Based Features.

Jia-Jie Mo,Chao Chen,Chao Zhang,Jun-Jian Zhou,Xiu Wang,Jian-Guo Zhang,Wen-Ling Li,Kai Zhang,Na-Jing Zhou,Yao Wang,Bao-Tian Zhao,Chang Liu,Wen-Han Hu

doi:10.3389/fnins.2018.01008

Abstract

Objective: To automatically detect focal cortical dysplasia (FCD) lesion by combining quantitative multimodal surface-based features with machine learning and to assess its clinical value.Methods: Neuroimaging data and clinical information for 74 participants (40 with histologically proven FCD type II) was retrospectively included. The morphology, intensity and function-based features characterizing FCD lesions were calculated vertex-wise on each cortical surface and fed to an artificial neural network. The classifier performance was quantitatively and qualitatively assessed by performing statistical analysis and conventional visual analysis.Results: The accuracy, sensitivity, specificity of the neural network classifier based on multimodal surface-based features were 70.5%, 70.0%, and 69.9%, respectively, which outperformed the unimodal classifier. There was no significant difference in the detection rate of FCD subtypes (Pearson’s Chi-Square = 0.001, p = 0.970). Cohen’s kappa score between automated detection outcomes and post-surgical resection region was 0.385 (considered as fair).Conclusion: Automated machine learning with multimodal surface features can provide objective and intelligent detection of FCD lesion in pre-surgical evaluation and can assist the surgical strategy. Furthermore, the optimal parameters, appropriate surface features and efficient algorithm are worth exploring.

Highlights

Focal cortical dysplasia (FCD) was intrinsically epileptogenic and was a significant cause of medically refractory epilepsy (Fauser, 2015)
Focal cortical dysplasia constituted a broad spectrum of histopathological and clinical features ranging from FCD type I to FCD type III (Blümcke et al, 2011)
18fluoro2-deoxy-d-glucose (18FDG) positron emission tomography– computed tomography (PET-CT) was performed to help with the localization of epileptogenic disturbances in metabolism, which may aid the identification of occult FCD that were missed on MRI

Summary

Introduction

Focal cortical dysplasia (FCD) was intrinsically epileptogenic and was a significant cause of medically refractory epilepsy (Fauser, 2015). Automated Detection of FCD lesions were focal, epilepsy surgery may be an option. Focal cortical dysplasia constituted a broad spectrum of histopathological and clinical features ranging from FCD type I (small or subtle in conventional magnetic resonance imaging [MRI]) to FCD type III (severe pathology with other associated epileptogenic lesions) (Blümcke et al, 2011). PET-CT often revealed focal hypometabolism in the FCD region and has been shown to have a diagnostic sensitivity of 78–83% in FCD detection (Chassoux et al, 2010; Yh et al, 2011). Despite enormous progress in neuroimaging techniques and computational methods, many lesions remain subtle to identify, as the sensitivity is approximately 70% of patients with FCD (Wang et al, 2013; Kini et al, 2016). In some cases, re-examination of MRI images indicates that lesions were missed during initial interpretation, and the pre-operative evaluation process was time-consuming and depends upon the experience of the interpreters, which may hinder the localization of the EZ and advancements of surgical treatments

Methods

Results

Discussion

Conclusion