The Role of Structure MRI in Diagnosing Autism

Abstract

This study proposes a Computer-Aided Diagnostic (CAD) system to diagnose subjects with autism spectrum disorder (ASD). The CAD system identifies morphological anomalies within the brain regions of ASD subjects. Cortical features are scored according to their contribution in diagnosing a subject to be ASD or typically developed (TD) based on a trained machine-learning (ML) model. This approach opens the hope for developing a new CAD system for early personalized diagnosis of ASD. We propose a framework to extract the cerebral cortex from structural MRI as well as identifying the altered areas in the cerebral cortex. This framework consists of the following five main steps: (i) extraction of cerebral cortex from structural MRI; (ii) cortical parcellation to a standard atlas; (iii) identifying ASD associated cortical markers; (iv) adjusting feature values according to sex and age; (v) building tailored neuro-atlases to identify ASD; and (vi) artificial neural networks (NN) are trained to classify ASD. The system is tested on the Autism Brain Imaging Data Exchange (ABIDE I) sites achieving an average balanced accuracy score of 97±2%. This paper demonstrates the ability to develop an objective CAD system using structure MRI and tailored neuro-atlases describing specific developmental patterns of the brain in autism.

Keywords: CAD; autism; classification; feature selection; hyper-parameter optimization; machine learning; structure MRI.

Figure 1

Overview of the proposed system starting from acquiring MRI volumes up to the diagnosis.

Figure 2

Morphological features extracted from brain surfaces by freesurfer.

Figure 3

Distribution of the $S_a$ values within different brain regions.

Figure 4

The flowchart of The RFECV algorithm.

Figure 5

Number of selected features vs. the maximum balanced accuracy score achieved using these features when applying RFECV using the four core classifiers, using the local model.

Figure 6

Personalized diagnosis.

Figure 7

The number of selected features vs. the balanced accuracy score when applying RFECV with different classifiers. The red vertical line labels the number of features corresponding to the maximum balanced accuracy score.

Figure 8

The highest testing balanced accuracy score ± standard deviation achieved by each of the optimized classifiers with applying RFECV with the core classifiers. The red dot labels the classifiers with the highest mean testing accuracy over the five-folds cross-validation.

Figure 9

The highest testing balanced accuracy score, plus or minus one standard deviation, achieved by each of the optimized classifiers without applying any feature selection algorithms.The red point labels the classifiers achieving the highest performance.

Figure 10

Visualization of the most frequent brain regions to be selected by RFECV+LG2.