Attention Deficit Hyperactivity Disorder Diagnosis using non-linear univariate and multivariate EEG measurements: a preliminary study

Functional connectivity and complexity analyses of resting-state fMRI in pre-adolescents demonstrating the behavioral symptoms of ADHD

Attention deficit hyperactivity disorder (ADHD) has been characterized by impairments among distributed functional brain networks, e.g., the frontoparietal network (FPN), default mode network (DMN), reward and motivation-related circuits (RMN), and salience network (SAL). In the current study, we evaluated the complexity and functional connectivity (FC) of resting state fMRI (rsfMRI) in pre-adolescents with the behavioral symptoms of ADHD, for pathology-relevant networks. We leveraged data from the Adolescent Brain and Cognitive Development (ABCD) Study. The final study sample included 63 children demonstrating the behavioral features of ADHD and 92 healthy control children matched on age, sex, and pubertal development status. For selected regions in the relevant networks, ANCOVA compared multiscale entropy (MSE) and FC between the groups. Finally, differences in the association between MSE and FC were evaluated. We found significantly reduced MSE along with increased FC within the FPN of pre-adolescents demonstrating the behavior symptoms of ADHD compared to matched healthy controls. Significant partial correlations between MSE and FC emerged in the FPN and RMN in the healthy controls however the association was absent in the participants demonstrating the behavior symptoms of ADHD. The current findings of complexity and FC in ADHD pathology support hypotheses of altered function of inhibitory control networks in ADHD.

ADHD-AID: Aiding Tool for Detecting Children's Attention Deficit Hyperactivity Disorder via EEG-Based Multi-Resolution Analysis and Feature Selection

The severe effects of attention deficit hyperactivity disorder (ADHD) among adolescents can be prevented by timely identification and prompt therapeutic intervention. Traditional diagnostic techniques are complicated and time-consuming because they are subjective-based assessments. Machine learning (ML) techniques can automate this process and prevent the limitations of manual evaluation. However, most of the ML-based models extract few features from a single domain. Furthermore, most ML-based studies have not examined the most effective electrode placement on the skull, which affects the identification process, while others have not employed feature selection approaches to reduce the feature space dimension and consequently the complexity of the training models. This study presents an ML-based tool for automatically identifying ADHD entitled "ADHD-AID". The present study uses several multi-resolution analysis techniques including variational mode decomposition, discrete wavelet transform, and empirical wavelet decomposition. ADHD-AID extracts thirty features from the time and time-frequency domains to identify ADHD, including nonlinear features, band-power features, entropy-based features, and statistical features. The present study also looks at the best EEG electrode placement for detecting ADHD. Additionally, it looks into the location combinations that have the most significant impact on identification accuracy. Additionally, it uses a variety of feature selection methods to choose those features that have the greatest influence on the diagnosis of ADHD, reducing the classification's complexity and training time. The results show that ADHD-AID has provided scores for accuracy, sensitivity, specificity, F1-score, and Mathew correlation coefficients of 0.991, 0.989, 0.992, 0.989, and 0.982, respectively, in identifying ADHD with 10-fold cross-validation. Also, the area under the curve has reached 0.9958. ADHD-AID's results are significantly higher than those of all earlier studies for the detection of ADHD in adolescents. These notable and trustworthy findings support the use of such an automated tool as a means of assistance for doctors in the prompt identification of ADHD in youngsters.

Investigation of frequency components embedded in EEG recordings underlying neuronal mechanism of cognitive control and attentional functions

Attentional cognitive control regulates the perception to enhance human behaviour. The current study examines the atltentional mechanisms in terms of time and frequency of EEG signals. The cognitive load is higher for processing local attentional stimulus, thereby demanding higher response time (RT) with low response accuracy (RA). On the other hand, the global attentional mechanisms broadly promote the perception while demanding a low cognitive load with faster RT and high RA. Attentional mechanisms refer to perceptual systems that afford and allocate the adaptive behaviours for prioritizing the processing of relevant stimuli based on the local and global features. The early sensory component of C1, which was associated with the local attentional mechanism, showed higher amplitudes than the global attentional mechanisms in parieto-occipital regions. Further, the local attentional mechanisms were also sustained in N2 and P3 components increasing higher amplitude in the left and right hemispheric sides of temporal regions (T7 and T8). Theta band frequency had shown higher power spectrum density (PSD) values while processing local attentional mechanisms. However, the significance of other frequency bands was noticeably minute. Hence, integrating the attentional mechanisms in terms of ERP and frequency signatures, a hybrid custom weight allocation model (CWAM) was built to assess and predict the contribution of insignificant channels to significant ones. The CWAM model was formulated based on the computational linear regression derivatives. All the derivatives are computationally derived the significant score while channelizing the hierarchical performance of each channel with respect to the frequent and deviant occurrences of global-local stimulus. This model enables us to configure the neural dynamicity of cognitive allocation of resources within the different locations of the human brain while processing the attentional stimulus. CWAM is reported to be the first model to evaluate the performance of the non-significant channels for enhancing the response of significant channels. The findings of the CWAM model suggest that the brain's performance may be determined by the underlying contribution of the non-significant channels.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11571-022-09888-x.

Functional Connectivity and Complexity Analyses of Resting-State fMRI in Pre-Adolescents with ADHD

Attention deficit hyperactivity disorder (ADHD) has been characterized by impairments among distributed functional brain networks, e.g., the frontoparietal network (FPN), default mode network (DMN), and reward and motivation-related circuits (RMN). In the current study, we evaluated the complexity and functional connectivity (FC) of resting state fMRI (rsfMRI) in pre-adolescents with ADHD for pathology-relevant networks. We leveraged data from the Adolescent Brain and Cognitive Development (ABCD) Study. The final study sample included 63 children with ADHD and 92 healthy control children matched on age, sex, and pubertal development status. For selected regions in relevant networks, ANCOVA compared multiscale entropy (MSE) and FC between the groups. Finally, differences in the association between MSE and FC were evaluated. We found significantly reduced MSE along with increased FC within the FPN of pre-adolescents with ADHD compared to matched healthy controls. Significant partial correlations between MSE and FC emerged in fewer regions in the participants with ADHD than in the controls. The observation of reduced entropy is consistent with existing literature using rsfMRI and other neuroimaging modalities. The current findings of complexity and FC in ADHD support hypotheses of altered function of inhibitory control networks in ADHD.

Multiscale entropy of ADHD children during resting state condition

This present study aims to investigate neural mechanisms underlying ADHD compared to healthy children through the analysis of the complexity and the variability of the EEG brain signal using multiscale entropy (MSE), EEG signal standard deviation (SDs), as well as the mean, standard deviation (SDp) and coefficient of variation (CV) of absolute spectral power (PSD). For this purpose, a sample of children diagnosed with attention-deficit/hyperactivity disorder (ADHD) between 6 and 17 years old were selected based on the number of trials and diagnostic agreement, 32 for the open-eyes (OE) experimental condition and 25 children for the close-eyes (CE) experimental condition. Healthy control subjects were age- and gender-matched with the ADHD group. The MSE and SDs of resting-state EEG activity were calculated on 34 time scales using a coarse-grained procedure. In addition, the PSD was averaged in delta, theta, alpha, and beta frequency bands, and its mean, SDp, and CV were calculated. The results show that the MSE changes with age during development, increases as the number of scales increases and has a higher amplitude in controls than in ADHD. The absolute PSD results show CV differences between subjects in low and beta frequency bands, with higher variability values in the ADHD group. All these results suggest an increased EEG variability and reduced complexity in ADHD compared to controls.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11571-022-09869-0.

A biomarker discovery framework for childhood anxiety

Introduction

Anxiety is the most common manifestation of psychopathology in youth, negatively affecting academic, social, and adaptive functioning and increasing risk for mental health problems into adulthood. Anxiety disorders are diagnosed only after clinical symptoms emerge, potentially missing opportunities to intervene during critical early prodromal periods. In this study, we used a new empirical approach to extracting nonlinear features of the electroencephalogram (EEG), with the goal of discovering differences in brain electrodynamics that distinguish children with anxiety disorders from healthy children. Additionally, we examined whether this approach could distinguish children with externalizing disorders from healthy children and children with anxiety.

Methods

We used a novel supervised tensor factorization method to extract latent factors from repeated multifrequency nonlinear EEG measures in a longitudinal sample of children assessed in infancy and at ages 3, 5, and 7 years of age. We first examined the validity of this method by showing that calendar age is highly correlated with latent EEG complexity factors (r = 0.77). We then computed latent factors separately for distinguishing children with anxiety disorders from healthy controls using a 5-fold cross validation scheme and similarly for distinguishing children with externalizing disorders from healthy controls.

Results

We found that latent factors derived from EEG recordings at age 7 years were required to distinguish children with an anxiety disorder from healthy controls; recordings from infancy, 3 years, or 5 years alone were insufficient. However, recordings from two (5, 7 years) or three (3, 5, 7 years) recordings gave much better results than 7 year recordings alone. Externalizing disorders could be detected using 3- and 5 years EEG data, also giving better results with two or three recordings than any single snapshot. Further, sex assigned at birth was an important covariate that improved accuracy for both disorder groups, and birthweight as a covariate modestly improved accuracy for externalizing disorders. Recordings from infant EEG did not contribute to the classification accuracy for either anxiety or externalizing disorders.

Conclusion

This study suggests that latent factors extracted from EEG recordings in childhood are promising candidate biomarkers for anxiety and for externalizing disorders if chosen at appropriate ages.

Sampling inequalities affect generalization of neuroimaging-based diagnostic classifiers in psychiatry

BACKGROUND

The development of machine learning models for aiding in the diagnosis of mental disorder is recognized as a significant breakthrough in the field of psychiatry. However, clinical practice of such models remains a challenge, with poor generalizability being a major limitation.

METHODS

Here, we conducted a pre-registered meta-research assessment on neuroimaging-based models in the psychiatric literature, quantitatively examining global and regional sampling issues over recent decades, from a view that has been relatively underexplored. A total of 476 studies (n = 118,137) were included in the current assessment. Based on these findings, we built a comprehensive 5-star rating system to quantitatively evaluate the quality of existing machine learning models for psychiatric diagnoses.

RESULTS

A global sampling inequality in these models was revealed quantitatively (sampling Gini coefficient (G) = 0.81, p < .01), varying across different countries (regions) (e.g., China, G = 0.47; the USA, G = 0.58; Germany, G = 0.78; the UK, G = 0.87). Furthermore, the severity of this sampling inequality was significantly predicted by national economic levels (β =  - 2.75, p < .001, R²_adj = 0.40; r =  - .84, 95% CI: - .41 to - .97), and was plausibly predictable for model performance, with higher sampling inequality for reporting higher classification accuracy. Further analyses showed that lack of independent testing (84.24% of models, 95% CI: 81.0-87.5%), improper cross-validation (51.68% of models, 95% CI: 47.2-56.2%), and poor technical transparency (87.8% of models, 95% CI: 84.9-90.8%)/availability (80.88% of models, 95% CI: 77.3-84.4%) are prevailing in current diagnostic classifiers despite improvements over time. Relating to these observations, model performances were found decreased in studies with independent cross-country sampling validations (all p < .001, BF₁₀ > 15). In light of this, we proposed a purpose-built quantitative assessment checklist, which demonstrated that the overall ratings of these models increased by publication year but were negatively associated with model performance.

CONCLUSIONS

Together, improving sampling economic equality and hence the quality of machine learning models may be a crucial facet to plausibly translating neuroimaging-based diagnostic classifiers into clinical practice.

The Future of Neurodiagnostics and Emergence of a New Science

Electroencepholography (EEG) is the oldest and original brain measurement technology. Since EEG was first used in clinical settings, the role of neurodiagnostic professionals has focused on two principal tasks that require specialized training. These include collecting the EEG recording, performed primarily by EEG Technologists, and interpreting the recording, generally done by physicians with proper specialization. Emerging technology appears to enable non-specialists to contribute to these tasks. Neurotechnologists may feel vulnerable to being displaced by new technology. A similar shift occurred in the last century when human "computers," employed to perform repetitive calculations needed to solve complex mathematics for the Manhattan and Apollo Projects, were displaced by new electronic computing machines. Many human "computers" seized on the opportunity created by the new computing technology to become the first computer programmers and create the new field of computer science. That transition offers insights for the future of neurodiagnostics. From its inception, neurodiagnostics has been an information processing discipline. Advances in dynamical systems theory, cognitive neuroscience, and biomedical informatics have created an opportunity for neurodiagnostic professionals to help create a new science of functional brain monitoring. A new generation of advanced neurodiagnostic professionals that bring together knowledge and skills in clinical neuroscience and biomedical informatics will benefit psychiatry, neurology, and precision healthcare, lead to preventive brain health through the lifespan, and lead the establishment of a new science of clinical neuroinformatics.

Evaluation of Risk of Bias in Neuroimaging-Based Artificial Intelligence Models for Psychiatric Diagnosis: A Systematic Review

IMPORTANCE

Neuroimaging-based artificial intelligence (AI) diagnostic models have proliferated in psychiatry. However, their clinical applicability and reporting quality (ie, feasibility) for clinical practice have not been systematically evaluated.

OBJECTIVE

To systematically assess the risk of bias (ROB) and reporting quality of neuroimaging-based AI models for psychiatric diagnosis.

EVIDENCE REVIEW

PubMed was searched for peer-reviewed, full-length articles published between January 1, 1990, and March 16, 2022. Studies aimed at developing or validating neuroimaging-based AI models for clinical diagnosis of psychiatric disorders were included. Reference lists were further searched for suitable original studies. Data extraction followed the CHARMS (Checklist for Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modeling Studies) and PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) guidelines. A closed-loop cross-sequential design was used for quality control. The PROBAST (Prediction Model Risk of Bias Assessment Tool) and modified CLEAR (Checklist for Evaluation of Image-Based Artificial Intelligence Reports) benchmarks were used to systematically evaluate ROB and reporting quality.

FINDINGS

A total of 517 studies presenting 555 AI models were included and evaluated. Of these models, 461 (83.1%; 95% CI, 80.0%-86.2%) were rated as having a high overall ROB based on the PROBAST. The ROB was particular high in the analysis domain, including inadequate sample size (398 of 555 models [71.7%; 95% CI, 68.0%-75.6%]), poor model performance examination (with 100% of models lacking calibration examination), and lack of handling data complexity (550 of 555 models [99.1%; 95% CI, 98.3%-99.9%]). None of the AI models was perceived to be applicable to clinical practices. Overall reporting completeness (ie, number of reported items/number of total items) for the AI models was 61.2% (95% CI, 60.6%-61.8%), and the completeness was poorest for the technical assessment domain with 39.9% (95% CI, 38.8%-41.1%).

CONCLUSIONS AND RELEVANCE

This systematic review found that the clinical applicability and feasibility of neuroimaging-based AI models for psychiatric diagnosis were challenged by a high ROB and poor reporting quality. Particularly in the analysis domain, ROB in AI diagnostic models should be addressed before clinical application.

The automated preprocessing pipe-line for the estimation of scale-wise entropy from EEG data (APPLESEED): Development and validation for use in pediatric populations

It is increasingly understood that moment-to-moment brain signal variability - traditionally modeled out of analyses as mere "noise" - serves a valuable functional role related to development, cognitive processing, and psychopathology. Multiscale entropy (MSE) - a measure of signal irregularity across temporal scales - is an increasingly popular analytic technique in human neuroscience calculated from time series such as electroencephalography (EEG) signals. MSE provides insight into the time-structure and (non)linearity of fluctuations in neural activity and network dynamics, capturing the brain's moment-to-moment complexity as it operates on multiple time scales. MSE is emerging as a powerful predictor of developmental processes and outcomes. However, differences in data preprocessing and MSE computation make it challenging to compare results across studies. Here, we (1) provide an introduction to MSE for developmental researchers, (2) demonstrate the effect of preprocessing procedures on scale-wise entropy estimates, and (3) establish a standardized EEG preprocessing and entropy estimation pipeline that adapts a critical modification to the original MSE algorithm, and generates reliable scale-wise entropy estimates capable of differentiating developmental stages and cognitive states. This novel pipeline - the Automated Preprocessing Pipe-Line for the Estimation of Scale-wise Entropy from EEG Data (APPLESEED) is fully automated, customizable, and freely available for download from https://github.com/mhpuglia/APPLESEED.

Impact of Different Styles of Online Course Videos on Students' Attention During the COVID-19 Pandemic

Background

The COVID-19 pandemic interfered with normal campus life, resulting in the need for the course to be conducted in an ideal online format. The purpose of this study is to analyze the impact of different styles of online political course videos on students' attention during the COVID-19 pandemic.

Methods

Four college students participated in this small sample study. They were required to conduct two sessions of the experiment, in which they were required to watch three different styles of course videos in each session. While watching the videos, their EEG signals were acquired. For the acquired EEG signals, the sample entropy (SampEn) features were extracted. On the other hand, Mayer's theories of multimedia technology provide guidance for teachers' online courses to enhance students' attention levels. The results of EEG signals analysis and Mayer's theories of multimedia technology were combined to compare and analyze the effects of three styles of instructional videos.

Results

Based on comparisons of the SampEn and Mayer's theories of multimedia technology analysis, the results suggest that online instruction in a style where the instructor and content appear on the screen at the same time and the instructor points out the location of the content as it is explained is more likely to elicit higher levels of students' attention.

Conclusions

During the COVID-19 pandemic, online instructional methods have an impact on students' classroom attention. It is essential for teachers to design online instructional methods based on students' classroom attention levels and some multimedia instructional techniques to improve students' learning efficiency.

Physical Exercise Effects on University Students’ Attention: An EEG Analysis Approach

Physically active breaks (AB) are currently being proposed as an interesting tool to improve students’ attention. Reviews and meta-analyses confirm their effect on attention, but also warned about the sparse evidence based on vigilance and university students. Therefore, this pilot study aimed to (a) determine the effects of AB in comparison with passive breaks on university students’ vigilance and (b) to validate an analysis model based on machine learning algorithms in conjunction with a multiparametric model based on electroencephalography (EEG) signal features. Through a counterbalanced within-subject experimental study, six university students (two female; mean age = 25.67, STD = 3.61) had their vigilance performances (i.e., response time in Psycho-Motor Vigilance Task) and EEG measured, before and after a lecture with an AB and another lecture with a passive break. A multiparametric model based on the spectral power, signal entropy and response time has been developed. Furthermore, this model, together with different machine learning algorithms, shows that for the taken signals there are significant differences after the AB lesson, implying an improvement in attention. These differences are most noticeable with the SVM with RBF kernel and ANNs with F1-score of 85% and 88%, respectively. In conclusion, results showed that students performed better on vigilance after the lecture with AB. Although limited, the evidence found could help researchers to be more accurate in their EEG analyses and lecturers and teachers to improve their students’ attentions in a proper way.

Automated classification of attention deficit hyperactivity disorder and conduct disorder using entropy features with ECG signals

BACKGROUND

The most prevalent neuropsychiatric disorder among children is attention deficit hyperactivity disorder (ADHD). ADHD presents with a high prevalence of comorbid disorders such as conduct disorder (CD). The lack of definitive confirmatory diagnostic tests for ADHD and CD make diagnosis challenging. The distinction between ADHD, ADHD + CD and CD is important as the course and treatment are different. Electrocardiography (ECG) signals may become altered in behavioral disorders due to brain-heart autonomic interactions. We have developed a software tool to categorize ADHD, ADHD + CD and CD automatically on ECG signals.

METHOD

ECG signals from participants were decomposed using empirical wavelet transform into various modes, from which entropy features were extracted. Robust ten-fold cross-validation with adaptive synthetic sampling (ADASYN) and z-score normalization were performed at each fold. Analysis of variance (ANOVA) technique was employed to determine the variability within the three classes, and obtained the most discriminatory features. Highly significant entropy features were then fed to classifiers.

RESULTS

Our model yielded the best classification results with the bagged tree classifier: 87.19%, 87.71% and 86.29% for accuracy, sensitivity and specificity, respectively.

CONCLUSION

The proposed expert system can potentially assist mental health professionals in the stratification of the three classes, for appropriate intervention using accessible ECG signals.

Autonomic responses to facial expression tasks in children with autism spectrum disorders: Cross-section study

BACKGROUND

The autonomic nervous system has an influence on emotions and behavior modulation, however, the relationship between autonomic modulation impairment and the autism spectrum disorder (ASD) is yet to be fully described.

AIMS

To evaluate the autonomic responses of children with and without ASD through the non-linear, and linear heart rate variability (HRV) measures, and assess the correlation between these responses, the severity and behavioral symptoms of autism.

METHODS AND PROCEDURES

27 children diagnosed with ASD (EG = experimental group) and 28 matching controls (CG = control group) were evaluated. The HRV was evaluated in 15 min sections at the following moments: I) Resting condition; II) During facial expression tasks; and III) Recovery. The severity and behavioral symptoms of autism were evaluated by the Childhood Autism Rating Scale (CARS) and Autistic Behaviors Checklist (ABC) scales.

OUTCOMES AND RESULTS

The facial expression tasks influenced the activity of the autonomic nervous system in both groups, however the EG experienced more autonomic changes. These changes were mostly evidenced by the non-linear indices. Also, the CARS and ABC scales showed significant correlations with HRV indices.

CONCLUSIONS AND IMPLICATIONS

Children with ASD presented an autonomic modulation impairment, mostly identified by the non-linear indices of HRV. Also, this autonomic impairment is associated with the severity and behavioral symptoms of autism.

A Modified Multivariable Complexity Measure Algorithm and Its Application for Identifying Mental Arithmetic Task

Properly measuring the complexity of time series is an important issue. The permutation entropy (PE) is a widely used as an effective complexity measurement algorithm, but it is not suitable for the complexity description of multi-dimensional data. In this paper, in order to better measure the complexity of multi-dimensional time series, we proposed a modified multivariable PE (MMPE) algorithm with principal component analysis (PCA) dimensionality reduction, which is a new multi-dimensional time series complexity measurement algorithm. The analysis results of different chaotic systems verify that MMPE is effective. Moreover, we applied it to the comlexity analysis of EEG data. It shows that the person during mental arithmetic task has higher complexity comparing with the state before mental arithmetic task. In addition, we also discussed the necessity of the PCA dimensionality reduction.

Measuring the effects of sleep on epileptogenicity with multifrequency entropy

OBJECTIVE

We demonstrate that multifrequency entropy gives insight into the relationship between epileptogenicity and sleep, and forms the basis for an improved measure of medical assessment of sleep impairment in epilepsy patients.

METHODS

Multifrequency entropy was computed from electroencephalography measurements taken from 31 children with Benign Epilepsy with Centrotemporal Spikes and 31 non-epileptic controls while awake and during sleep. Values were compared in the epileptic zone and away from the epileptic zone in various sleep stages.

RESULTS

We find that (I) in lower frequencies, multifrequency entropy decreases during non-rapid eye movement sleep stages when compared with wakefulness in a general population of pediatric patients, (II) patients with Benign Epilepsy with Centrotemporal Spikes had lower multifrequency entropy across stages of sleep and wakefulness, and (III) the epileptic regions of the brain exhibit lower multifrequency entropy patterns than the rest of the brain in epilepsy patients.

CONCLUSIONS

Our results show that multifrequency entropy decreases during sleep, particularly sleep stage 2, confirming, in a pediatric population, an association between sleep, lower multifrequency entropy, and increased likelihood of seizure.

SIGNIFICANCE

We observed a correlation between lowered multifrequency entropy and increased epileptogenicity that lays preliminary groundwork for the detection of a digital biomarker for epileptogenicity.

Frontal EEG-Based Multi-Level Attention States Recognition Using Dynamical Complexity and Extreme Gradient Boosting

Measuring and identifying the specific level of sustained attention during continuous tasks is essential in many applications, especially for avoiding the terrible consequences caused by reduced attention of people with special tasks. To this end, we recorded EEG signals from 42 subjects during the performance of a sustained attention task and obtained resting state and three levels of attentional states using the calibrated response time. EEG-based dynamical complexity features and Extreme Gradient Boosting (XGBoost) classifier were proposed as the classification model, Complexity-XGBoost, to distinguish multi-level attention states with improved accuracy. The maximum average accuracy of Complexity-XGBoost were 81.39 ± 1.47% for four attention levels, 80.42 ± 0.84% for three attention levels, and 95.36 ± 2.31% for two attention levels in 5-fold cross-validation. The proposed method is compared with other models of traditional EEG features and different classification algorithms, the results confirmed the effectiveness of the proposed method. We also found that the frontal EEG dynamical complexity measures were related to the changing process of response during sustained attention task. The proposed dynamical complexity approach could be helpful to recognize attention status during important tasks to improve safety and efficiency, and be useful for further brain-computer interaction research in clinical research or daily practice, such as the cognitive assessment or neural feedback treatment of individuals with attention deficit hyperactivity disorders, Alzheimer’s disease, and other diseases which affect the sustained attention function.

Automated detection of conduct disorder and attention deficit hyperactivity disorder using decomposition and nonlinear techniques with EEG signals

BACKGROUND AND OBJECTIVES

Attention deficit hyperactivity disorder (ADHD) is often presented with conduct disorder (CD). There is currently no objective laboratory test or diagnostic method to discern between ADHD and CD, and diagnosis is further made difficult as ADHD is a common neuro-developmental disorder often presenting with other co-morbid difficulties; and in particular with conduct disorder which has a high degree of associated behavioural challenges. A novel automated system (AS) is proposed as a convenient supplementary tool to support clinicians in their diagnostic decisions. To the best of our knowledge, we are the first group to develop an automated classification system to classify ADHD, CD and ADHD+CD classes using brain signals.

METHODS

The empirical mode decomposition (EMD) and discrete wavelet transform (DWT) methods were employed to decompose the electroencephalogram (EEG) signals. Autoregressive modelling coefficients and relative wavelet energy were then computed on the signals. Various nonlinear features were extracted from the decomposed coefficients. Adaptive synthetic sampling (ADASYN) was then employed to balance the dataset. The significant features were selected using sequential forward selection method. The highly discriminatory features were subsequently fed to an array of classifiers.

RESULTS

The highest accuracy of 97.88% was achieved with the K-Nearest Neighbour (KNN) classifier. The proposed system was developed using ten-fold validation strategy on EEG data from 123 children. To the best of our knowledge this is the first study to develop an AS for the classification of ADHD, CD and ADHD+CD classes using EEG signals.

POTENTIAL APPLICATION

Our AS can potentially be used as a web-based application with cloud system to aid the clinical diagnosis of ADHD and/or CD, thus supporting faster and accurate treatment for the children. It is important to note that testing with larger data is required before the AS can be employed for clinical applications.