An EEG based Channel Optimized Classification Approach for Autism Spectrum Disorder

Abstract

Spectrum Disorder (ASD) is a neurodevelopmental condition which affects a persons cognition and behaviour. It is a lifelong condition which cannot be cured completely using any intervention to date. However, early diagnosis and follow-up treatments have a major impact on autistic people. Unfortunately, the current diagnostic practices, which are subjective and behaviour dependent, delay the diagnosis at an early age and makes it harder to distinguish autism from other developmental disorders. Several works of literature explore the possible behaviour-independent measures to diagnose ASD. Abnormalities in EEG can be used as reliable biomarkers to diagnose ASD. This work presents a low-cost and straightforward diagnostic approach to classify ASD based on EEG signal processing and learning models. Possibilities to use a minimum number of EEG channels have been explored. Statistical features are extracted from noise filtered EEG data before and after Discrete Wavelet Transform. Relevant features and EEG channels were selected using correlation-based feature selection. Several learning models and feature vectors have been studied and possibilities to use the minimum number of EEG channels have also been explored. Using Random Forest and Correlation-based Feature Selection, an accuracy level of 93% was obtained.

Autism Spectrum Disorder Diagnostic System Using HOS Bispectrum with EEG Signals

“…From Table 3, it is apparent that nonlinear features have been prevalently used to diagnose AD [49,50,54,55,57]. Additionally, SVM classifiers have also been commonly employed to classify EEG signals for the detection of ASD [52,54,56,[58][59][60] similar to our study. Although a classification study was done, lower accuracies were achieved in the following studies: [52,54,57,59,60] as compared to ours.…”

“…Additionally, SVM classifiers have also been commonly employed to classify EEG signals for the detection of ASD [52,54,56,[58][59][60] similar to our study. Although a classification study was done, lower accuracies were achieved in the following studies: [52,54,57,59,60] as compared to ours. Although higher classification accuracies of 100% [58] and 99.71% [47] were achieved in these particular two studies as compared to our study, smaller data sizes were used for training in both studies.…”

“…The features were then input to various classifiers. A highest accuracy of 93% was achieved with the random forest classifier, using k-fold validation [59].…”

EEG-based Processing and Classification Methodologies for Autism Spectrum Disorder: A Review

“…Signal decomposition using DWT is shown in Fig. 2 (Bosl et al, 2018) O X ASD classification using EEG and eye movement (Thapaliya et al, 2018) X X Classifying ASD using MS-ROM/I-FAST algorithm (Grossi et al, 2017) X ASD diagnosis using DWT, Shannon entropy and ANN (Djemal et al, 2017) O X X X Wavelet-based ASD classification (Cheong et al, 2015) O X X ASD diagnosis utilizing brain connectivity (Jamal et al, 2014) X X O Fuzzy synchronization likelihood methodology for ASD diagnosis (Ahmadlou et al, 2012a) O ASD diagnosis based on improved visibility graph fractality (Ahmadlou et al, 2012b) O X EEG as a biomarker for distinguishing ASD children (Bosl et al, 2011) X Classification of ASD using fractal dimensions (Ahmadlou et al, 2010) O X Frequency 3D mapping and interchannel stability of EEG as indicators towards ASD diagnosis (Abdulhay et al, 2017) X X O O Diagnosing ASD utilizing EEG spectral coherence (Duffy and Als, 2012) X X X X ASDGenus: channel optimised classification using EEG (Haputhanthri et al, 2019) X O X…”

“…Improves visibility graph fractality Needs large dataset, evaluated 95.5% 19 2012b) based ASD classification, using only the holdout method. Better noise robustness (Haputhanthri et al, 2019) Uses a smaller number of channels Need to be evaluated on a larger 93.33% 5 ensuring simplicity and channel optimisation population to ensure statistical significance of the results (Bosl et al, Uses modified multiscale entropy as a Needs many EEG channels 90% 64 2011) biomarker for ASD, Age is between 6 and 24 months, classify based on gender, age groups. (Ahmadlou et al, Classification approach-based Needs large dataset, evaluated 90% 19 2010) on fractal dimensions using only the holdout method.…”