Machine learning approaches for electroencephalography and magnetoencephalography analyses in autism spectrum disorder: A systematic review

Identification of autism spectrum disorder based on electroencephalography: A systematic review

Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder characterized by difficulties in social communication and repetitive and stereotyped behaviors. According to the World Health Organization, about 1 in 100 children worldwide has autism. With the global prevalence of ASD, timely and accurate diagnosis has been essential in enhancing the intervention effectiveness for ASD children. Traditional ASD diagnostic methods rely on clinical observations and behavioral assessment, with the disadvantages of time-consuming and lack of objective biological indicators. Therefore, automated diagnostic methods based on machine learning and deep learning technologies have emerged and become significant since they can achieve more objective, efficient, and accurate ASD diagnosis. Electroencephalography (EEG) is an electrophysiological monitoring method that records changes in brain spontaneous potential activity, which is of great significance for identifying ASD children. By analyzing EEG data, it is possible to detect abnormal synchronous neuronal activity of ASD children. This paper gives a comprehensive review of the EEG-based ASD identification using traditional machine learning methods and deep learning approaches, including their merits and potential pitfalls. Additionally, it highlights the challenges and the opportunities ahead in search of more effective and efficient methods to automatically diagnose autism based on EEG signals, which aims to facilitate automated ASD identification.

Autism Spectrum Disorder: Brain Areas Involved, Neurobiological Mechanisms, Diagnoses and Therapies

Autism spectrum disorder (ASD), affecting over 2% of the pre-school children population, includes an important fraction of the conditions accounting for the heterogeneity of autism. The disease was discovered 75 years ago, and the present review, based on critical evaluations of the recognized ASD studies from the beginning of 1990, has been further developed by the comparative analyses of the research and clinical reports, which have grown progressively in recent years up to late 2023. The tools necessary for the identification of the ASD disease and its related clinical pathologies are genetic and epigenetic mutations affected by the specific interaction with transcription factors and chromatin remodeling processes occurring within specific complexes of brain neurons. Most often, the ensuing effects induce the inhibition/excitation of synaptic structures sustained primarily, at dendritic fibers, by alterations of flat and spine response sites. These effects are relevant because synapses, established by specific interactions of neurons with glial cells, operate as early and key targets of ASD. The pathology of children is often suspected by parents and communities and then confirmed by ensuing experiences. The final diagnoses of children and mature patients are then completed by the combination of neuropsychological (cognitive) tests and electro-/magneto-encephalography studies developed in specialized centers. ASD comorbidities, induced by processes such as anxieties, depressions, hyperactivities, and sleep defects, interact with and reinforce other brain diseases, especially schizophrenia. Advanced therapies, prescribed to children and adult patients for the control of ASD symptoms and disease, are based on the combination of well-known brain drugs with classical tools of neurologic and psychiatric practice. Overall, this review reports and discusses the advanced knowledge about the biological and medical properties of ASD.

Hybrid deep transfer learning-based early diagnosis of autism spectrum disorder using scalogram representation of electroencephalography signals

Early diagnosis of autism spectrum disorder (ASD) plays an important role in the rehabilitation of the patient. This goal necessitates higher-level pattern representation and a strong modeling approach. The proposed approach applies scalogram images of electroencephalography signals for the first purpose and a two-level deep learning architecture for better classification. Scalogram images embed both the temporal and spectral information of the signal. On the other hand, the hybrid deep learning hierarchy of convolutional neural network followed by long short-term memory models both spatial and temporal information of the scalogram image. The approach is evaluated on a dataset of 34 ASD samples and 11 normal cases in without-voice and with-voice conditions. To validate the early diagnosis hypothesis, signals from children older than 5 years are used as the training set, and signals from younger subjects are used as the validation set. The proposed method achieves excellent performance of 99.50% and 98.43% for automatically detecting ASD with and without voice, respectively. This classification performance is higher than most recent reported approaches, and the results show the effectiveness of the approach in early diagnosis of ASD and demonstrate the auditory impact on the diagnosis of autism.

Classification of autism spectrum disorder using electroencephalography in Chinese children: a cross-sectional retrospective study

Autism spectrum disorder (ASD) is a complex neurodevelopmental condition characterized by diverse clinical features. EEG biomarkers such as spectral power and functional connectivity have emerged as potential tools for enhancing early diagnosis and understanding of the neural processes underlying ASD. However, existing studies yield conflicting results, necessitating a comprehensive, data-driven analysis. We conducted a retrospective cross-sectional study involving 246 children with ASD and 42 control children. EEG was collected, and diverse EEG features, including spectral power and spectral coherence were extracted. Statistical inference methods, coupled with machine learning models, were employed to identify differences in EEG features between ASD and control groups and develop classification models for diagnostic purposes. Our analysis revealed statistically significant differences in spectral coherence, particularly in gamma and beta frequency bands, indicating elevated long range functional connectivity between frontal and parietal regions in the ASD group. Machine learning models achieved modest classification performance of ROC-AUC at 0.65. While machine learning approaches offer some discriminative power classifying individuals with ASD from controls, they also indicate the need for further refinement.