A hybrid method to select morphometric features using tensor completion and F-score rank for gifted children identification

Abstract

Gifted children are able to learn in a more advanced way than others, probably due to neurophysiological differences in the communication efficiency in neural pathways. Topological features contribute to understanding the correlation between the brain structure and intelligence. Despite decades of neuroscience research using MRI, methods based on brain region connectivity patterns are limited by MRI artifacts, which therefore leads to revisiting MRI morphometric features, with the aim of using them to directly identify gifted children instead of using brain connectivity. However, the small, high-dimensional morphometric feature dataset with outliers makes the task of finding good classification models challenging. To this end, a hybrid method is proposed that combines tensor completion and feature selection methods to handle outliers and then select the discriminative features. The proposed method can achieve a classification accuracy of 93.1%, higher than other existing algorithms, which is thus suitable for the small MRI datasets with outliers in supervised classification scenarios.

Graph Empirical Mode Decomposition-Based Data Augmentation Applied to Gifted Children MRI Analysis

“…In this section, we evaluate the performance of the GEMD using the procedures described in Zhang et al ( 2021 ). In summary, Zhang et al proposed an outlier correction on the morphometric features based on the STDC algorithm (Chen et al, 2013 ) and explored several feature selection methods to classify MRI data from controls and gifted groups.…”

“…According to Zhang et al ( 2021 ), the NONE feature selection method used all the features in the raw feature matrix. The VON feature selection method used only the regions belonging to types 2 and 3 of the von Economo atlas (van den Heuvel et al, 2015 ), which corresponds to the associative areas of the brain.…”

“…Finally, the combination of VON and FS will lead the VFS feature selection method, in which only type 2 and 3 regions are considered when calculating the FS value for morphometric features. Two traditional machine learning methods, KNN and SVM, were used as classifiers (Zhang et al, 2021 ), with leave-one-out as a cross-validation strategy.…”

“…We train the BrainNetCNN for the classification task, showing that a well-trained classification model can increase the classification accuracy from 55.7 to 78% when using artificial data. Moreover, in Zhang et al ( 2021 ), a hybrid selection method of morphological features and brain regions on the same gifted children dataset was derived. They used a completion method, simultaneous tensor decomposition, and completion (STDC), for outlier correction.…”