Classification of Bipolar Disorder and Schizophrenia Using Steady-State Visual Evoked Potential Based Features

Han

et al. 2018

Sensors

Self Cite

Brain-computer interface (BCI) studies based on electroencephalography (EEG) measured around the ears (ear-EEGs) have mostly used exogenous paradigms involving brain activity evoked by external stimuli. The objective of this study is to investigate the feasibility of ear-EEGs for development of an endogenous BCI system that uses self-modulated brain activity. We performed preliminary and main experiments where EEGs were measured on the scalp and behind the ears to check the reliability of ear-EEGs as compared to scalp-EEGs. In the preliminary and main experiments, subjects performed eyes-open and eyes-closed tasks, and they performed mental arithmetic (MA) and light cognitive (LC) tasks, respectively. For data analysis, the brain area was divided into four regions of interest (ROIs) (i.e., frontal, central, occipital, and ear area). The preliminary experiment showed that the degree of alpha activity increase of the ear area with eyes closed is comparable to those of other ROIs (occipital > ear > central > frontal). In the main experiment, similar event-related (de)synchronization (ERD/ERS) patterns were observed between the four ROIs during MA and LC, and all ROIs showed the mean classification accuracies above 70% required for effective binary communication (MA vs. LC) (occipital = ear = central = frontal). From the results, we demonstrated that ear-EEG can be used to develop an endogenous BCI system based on cognitive tasks without external stimuli, which allows the usability of ear-EEGs to be extended.

Section: Discussionmentioning

confidence: 99%

On the Feasibility of Using an Ear-EEG to Develop an Endogenous Brain-Computer Interface

Han

et al. 2018

Sensors

Self Cite

“…Microstate analysis can be a very useful method for exploring the brain network functions of patients with mental illness (35). Scholars are currently investigating the potential applications of electroencephalograms (EEGs), and the characteristics of scalp EEG are commonly applied to distinguish between BD and SCH (36).…”

Section: Introductionmentioning

confidence: 99%

Electroencephalographic Microstates in Schizophrenia and Bipolar Disorder

2021

Schizophrenia (SCH) and bipolar disorder (BD) are characterized by many types of symptoms, damaged cognitive function, and abnormal brain connections. The microstates are considered to be the cornerstones of the mental states shown in EEG data. In our study, we investigated the use of microstates as biomarkers to distinguish patients with bipolar disorder from those with schizophrenia by analyzing EEG data measured in an eyes-closed resting state. The purpose of this article is to provide an electron directional physiological explanation for the observed brain dysfunction of schizophrenia and bipolar disorder patients.Methods: We used microstate resting EEG data to explore group differences in the duration, coverage, occurrence, and transition probability of 4 microstate maps among 20 SCH patients, 26 BD patients, and 35 healthy controls (HCs).Results: Microstate analysis revealed 4 microstates (A–D) in global clustering across SCH patients, BD patients, and HCs. The samples were chosen to be matched. We found the greater presence of microstate B in BD patients, and the less presence of microstate class A and B, the greater presence of microstate class C, and less presence of D in SCH patients. Besides, a greater frequent switching between microstates A and B and between microstates B and A in BD patients than in SCH patients and HCs and less frequent switching between microstates C and D and between microstates D and C in BD patients compared with SCH patients.Conclusion: We found abnormal features of microstate A, B in BD patients and abnormal features of microstate A, B, C, and D in SCH patients. These features may indicate the potential abnormalities of SCH patients and BD patients in distributing neural resources and influencing opportune transitions between different states of activity.

“…The used features were as follows: sensor-level – each of the sixty-two P300 amplitudes and latencies (total 124); source-level – CSD in different brain regions showing significantly different between groups. In order to decrease the computational cost and avoid the overfitting by the use of large numbers of features, feature selection based on Fisher score(Alimardani et al, 2018b; Gu et al, 2012) was used. The higher Fisher score for each feature represents the better discrimination capability between the two groups.…”

Section: Methodsmentioning

confidence: 99%

“…The features with relatively higher Fisher scores were selected for the classification, with the number of features ranging from 1 to 20 (Shim et al, 2016). The classification accuracy was evaluated using a two-class linear support vector machine (SVM) classifier (Alimardani et al, 2018a; Orru et al, 2012) with a leave-one-out cross-validation (LOOCV) method for each feature set (Shim et al, 2016). To compute LOOCV, the only one subject was used for validating the model and remaining subjects (N-1) were used for training of the model (Al-Kaysi et al, 2017; Wang et al, 2015).…”

Section: Methodsmentioning

confidence: 99%

Machine-learning-based classification between post-traumatic stress disorder and major depressive disorder using P300 features

Shim

Jin

et al. 2019

NeuroImage: Clinical

BackgroundThe development of optimal classification criteria for specific mental disorders which share similar symptoms is an important issue for precise diagnosis. We investigated whether P300 features in both sensor-level and source-level could be effectively used to classify post-traumatic stress disorder (PTSD) and major depressive disorder (MDD).MethodEEG signals were recorded from fifty-one PTSD patients, 67 MDD patients, and 39 healthy controls (HCs) while performing an auditory oddball task. Amplitude and latency of P300 were evaluated, and the current source analysis of P300 components was conducted using sLORETA. Finally, we classified two groups using machine-learning methods with both sensor- and source-level features. Moreover, we checked the comorbidity effects using the same approaches (PTSD-mono diagnosis (PTSDm, n = 28) and PTSD-comorbid diagnosis (PTSDc, n = 23)).ResultsPTSD showed significantly reduced P300 amplitudes and prolonged latency compared to HCs and MDD. Moreover, PTSD showed significantly reduced source activities, and the source activities were significantly correlated with symptoms of depression and anxiety. Also, the best classification accuracy at each pair was as follows: 80.00% (PTSD-HCs), 67.92% (MDD-HCs), 70.34% (PTSD-MDD), 82.09% (PTSDm-HCs), 71.58% (PTSDm-MDD), 82.56% (PTSDc-HCs), and 76.67% (PTSDc- MDD).ConclusionSince abnormal P300 reflects pathophysiological characteristics of PTSD, PTSD patients were well-discriminated from MDD and HCs when using P300 features. Thus, altered P300 characteristics in both sensor- and source-level may be useful biomarkers to diagnosis PTSD.