Background/Objectives: Alzheimer’s disease (AD) is a progressive neurodegenerative disorder advancing through subjective cognitive decline (SCD), mild cognitive impairment (MCI), and dementia, making early diagnosis crucial. Electroencephalography (EEG) is a non-invasive, cost-effective alternative to advanced neuroimaging for detecting early neural changes. While most studies focus on resting-state EEG or handcrafted features with traditional machine learning, deep learning (DL) offers a promising tool for automated EEG analysis. This study classified the AD spectrum (SCD, MCI, AD) using EEG recorded during resting-state and task-based conditions. Specifically, EEG was recorded during a simple yes/no question-answering task, mimicking everyday cognitive activities, and was explored. We hypothesized that brain activity during tasks involving listening, comprehension, and response execution provides diagnostic insights. Methods: We collected 1 min of resting-state EEG and approximately 3 min of task-based EEG from 20, 28, and 10 participants with SCD, MCI, and AD, respectively. Task data included response accuracy and reaction time. After minimal preprocessing, two DL models, attention long short-term memory and EEGConformer, were used for binary (e.g., SCD vs. MCI) and three-class (SCD, MCI, AD) classification. Results: Task-based EEG outperformed resting-state EEG, with a 5–15% improvement in accuracy. The area under the curve (AUC) results consistently demonstrated superior classification performance for task-based EEG compared to resting-state EEG across all group distinctions. No significant performance difference was observed between the two DL models. Conclusions: We proposed a cognitive task-based approach for early AD spectrum diagnosis via EEG, offering greater accuracy by leveraging advanced DL models.
Neurodegenerative change in the central nervous system has been suggested as one of the pathophysiological mechanisms of autonomic nervous system dysfunction in Parkinson’s disease (PD). We analyzed gray matter (GM) volume changes and clinical parameters in patients with PD to investigate any involvement in the brain structures responsible for autonomic control in patients with PD having orthostatic hypotension (OH). Voxel-based morphometry was applied to compare regional GM volumes between PD patients with and without OH. Multivariate logistic regression analysis using a hierarchical model was carried out to identify clinical factors independently contributing to the regional GM volume changes in PD patients with OH. The Sobel test was used to analyze mediation effects between the independent contributing factors to the GM volume changes. PD patients with OH had more severe autonomic dysfunction and reduction in volume in the right inferior temporal cortex than those without OH. The right inferior temporal volume was positively correlated with the Qualitative Scoring MMSE Pentagon Test (QSPT) score, reflecting visuospatial/visuoperceptual function, and negatively correlated with the Composite Autonomic Severity Score (CASS). The CASS and QSPT scores were found to be factors independently contributing to regional volume changes in the right inferior temporal cortex. The QSPT score was identified as a mediator in which regional GM volume predicts the CASS. Our findings suggest that a decrease in the visuospatial/visuoperceptual process may be involved in the presentation of autonomic nervous system dysfunction in PD patients.
We aimed to compare network properties between focal-onset nonconvulsive status epilepticus (NCSE) and toxic/metabolic encephalopathy (TME) during periods of periodic discharge using graph theoretical analysis, and to evaluate the applicability of graph measures as markers for the differential diagnosis between focal-onset NCSE and TME, using machine learning algorithms. Electroencephalography (EEG) data from 50 focal-onset NCSE and 44 TMEs were analyzed. Epochs with nonictal periodic discharges were selected, and the coherence in each frequency band was analyzed. Graph theoretical analysis was performed to compare brain network properties between the groups. Eight different traditional machine learning methods were implemented to evaluate the utility of graph theoretical measures as input features to discriminate between the two conditions. The average degree (in delta, alpha, beta, and gamma bands), strength (in delta band), global efficiency (in delta and alpha bands), local efficiency (in delta band), clustering coefficient (in delta band), and transitivity (in delta band) were higher in TME than in NCSE. TME showed lower modularity (in delta band) and assortativity (in alpha, beta, and gamma bands) than NCSE. Machine learning algorithms based on EEG global graph measures classified NCSE and TME with high accuracy, and gradient boosting was the most accurate classification model with an area under the receiver operating characteristics curve of 0.904. Our findings on differences in network properties may provide novel insights that graph measures reflecting the network properties could be quantitative markers for the differential diagnosis between focal-onset NCSE and TME.
We compared neural activities and network properties between the antihistamine-induced seizures (AIS) and seizure-free groups, with the hypothesis that patients with AIS might have inherently increased neural activities and network properties that are easily synchronized. Resting-state electroencephalography (EEG) data were collected from 27 AIS patients and 30 healthy adults who had never had a seizure. Power spectral density analysis was used to compare neural activities in each localized region. Functional connectivity (FC) was measured using coherence, and graph theoretical analyses were performed to compare network properties between the groups. Machine learning algorithms were applied using measurements found to be different between the groups in the EEG analyses as input features. Compared with the seizure-free group, the AIS group showed a higher spectral power in the entire regions of the delta, theta, and beta bands, as well as in the frontal areas of the alpha band. The AIS group had a higher overall FC strength, as well as a shorter characteristic path length in the theta band and higher global efficiency, local efficiency, and clustering coefficient in the beta band than the seizure-free group. The Support Vector Machine, k-Nearest Neighbor, and Random Forest models distinguished the AIS group from the seizure-free group with a high accuracy of more than 99%. The AIS group had seizure susceptibility considering both regional neural activities and functional network properties. Our findings provide insights into the underlying pathophysiological mechanisms of AIS and may be useful for the differential diagnosis of new-onset seizures in the clinical setting.
About Me
Department of Neurology, Korea university medicine, 73, Goryeodae-ro, Seongbuk-gu, Seoul, Republic of Korea, 02841
Department of Neurology, Korea university medicine, 73, Goryeodae-ro, Seongbuk-gu, Seoul, Republic of Korea, 02841