인문학
사회과학
자연과학
공학
의약학
농수해양학
예술체육학
복합학
DBpia AI
AI 에이전트
AI 아이디어
AI 리더
AI 뷰어
Citeasy
크롬 서지관리 다운로드
지원사업
학술연구/단체지원/교육 등 연구자 활동을 지속하도록 DBpia가 지원하고 있어요.
커뮤니티
연구자들이 자신의 연구와 전문성을 널리 알리고, 새로운 협력의 기회를 만들 수 있는 네트워킹 공간이에요.
논문 기본 정보
- 저자정보
피인용 0
초록·키워드
Electroencephalogram (EEG) is widely used for clinical diagnosis and brain-computer interface (BCI) applications. In Alzheimer's disease (AD), disrupted functional connec-tivity between brain regions is a well-established biomarker. However, conventional EEG analysis methods often overlook inter-channel dependencies and are susceptible to noise due to the inherently low signal-to-noise ratio (SNR) of EEG signals, which limits robust feature extraction.
To address these limitations, we propose AlCheNet, a hybrid graph-based neural network that learns frequency-specific features and inter-channel dependencies in EEG data. The model first applies one-dimensional convolutions to extract frequency-specific features, then constructs a graph where each node corresponds to an EEG channel. A graph-based neural network is then employed to learn inter-channel relationships and denoise the signal by enforcing inter-channel smoothness. To avoid over-smoothing and preserve disease-relevant regional patterns, we used pathology-informed graph to guide adjacency matrix construction. We evaluated AlCheNet on a clinical EEG dataset comprising patients with AD, frontotemporal dementia, and healthy controls. AlCheNet, built upon the ChebNet architecture, outperformed a comparative ChebNet-based model by 1.82% in accuracy and 4.39% in F1-score, demonstrating superior representation learning capability
To address these limitations, we propose AlCheNet, a hybrid graph-based neural network that learns frequency-specific features and inter-channel dependencies in EEG data. The model first applies one-dimensional convolutions to extract frequency-specific features, then constructs a graph where each node corresponds to an EEG channel. A graph-based neural network is then employed to learn inter-channel relationships and denoise the signal by enforcing inter-channel smoothness. To avoid over-smoothing and preserve disease-relevant regional patterns, we used pathology-informed graph to guide adjacency matrix construction. We evaluated AlCheNet on a clinical EEG dataset comprising patients with AD, frontotemporal dementia, and healthy controls. AlCheNet, built upon the ChebNet architecture, outperformed a comparative ChebNet-based model by 1.82% in accuracy and 4.39% in F1-score, demonstrating superior representation learning capability
본문·목차
인공지능 문자 인식 모델을 통해 추출된 텍스트로, 일부 오타나 오류가 포함될 수 있으나 지속적으로 개선 중입니다.
오류를 발견하셨다면 해당 부분을 드래그한 후 ' 를 통해 신고해주세요.
오류를 발견하셨다면 해당 부분을 드래그한 후 ' 를 통해 신고해주세요.
최근 본 자료 전체보기
UCI(KEPA) : I410-151-25-02-093698937
