A deep learning model for the detection of various dementia and MCI pathologies based on resting-state electroencephalography data: A retrospective multicentre study

• Published in: Neural networks Vol. 171; pp. 242 - 250
• Main Authors: Watanabe, Yusuke, Miyazaki, Yuki, Hata, Masahiro, Fukuma, Ryohei, Aoki, Yasunori, Kazui, Hiroaki, Araki, Toshihiko, Taomoto, Daiki, Satake, Yuto, Suehiro, Takashi, Sato, Shunsuke, Kanemoto, Hideki, Yoshiyama, Kenji, Ishii, Ryouhei, Harada, Tatsuya, Kishima, Haruhiko, Ikeda, Manabu, Yanagisawa, Takufumi
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 01.03.2024
• Subjects: Aged; Alzheimer Disease - diagnosis; Alzheimer's disease; Cognitive Dysfunction - diagnosis; Deep Learning; Dementia screening; Dementia with Lewy bodies; Electroencephalography; Humans; Idiopathic normal-pressure hydrocephalus; Lewy Body Disease - diagnosis; Mild cognitive impairment; Multiple centres; Dementia screening; Multiple centres; Idiopathic normal-pressure hydrocephalus; Mild cognitive impairment; Alzheimer's disease; Dementia with Lewy bodies
• ISSN: 0893-6080; 1879-2782
• DOI: 10.1016/j.neunet.2023.12.009

•Convolutional neural network for electroencephalography (EEG) data was developed.•The system accurately classified patients with dementia diseases and healthy controls.•Prediction scores were consistent across institutions that were novel for the system.•The underlying pathology of mild cognitive impairment (MCI) was identified as well.•The existence of common EEG features among dementia diseases and MCI was implied. Dementia and mild cognitive impairment (MCI) represent significant health challenges in an aging population. As the search for noninvasive, precise and accessible diagnostic methods continues, the efficacy of electroencephalography (EEG) combined with deep convolutional neural networks (DCNNs) in varied clinical settings remains unverified, particularly for pathologies underlying MCI such as Alzheimer's disease (AD), dementia with Lewy bodies (DLB) and idiopathic normal-pressure hydrocephalus (iNPH). Addressing this gap, our study evaluates the generalizability of a DCNN trained on EEG data from a single hospital (Hospital #1). For data from Hospital #1, the DCNN achieved a balanced accuracy (bACC) of 0.927 in classifying individuals as healthy (n = 69) or as having AD, DLB, or iNPH (n = 188). The model demonstrated robustness across institutions, maintaining bACCs of 0.805 for data from Hospital #2 (n = 73) and 0.920 at Hospital #3 (n = 139). Additionally, the model could differentiate AD, DLB, and iNPH cases with bACCs of 0.572 for data from Hospital #1 (n = 188), 0.619 for Hospital #2 (n = 70), and 0.508 for Hospital #3 (n = 139). Notably, it also identified MCI pathologies with a bACC of 0.715 for Hospital #1 (n = 83), despite being trained on overt dementia cases instead of MCI cases. These outcomes confirm the DCNN's adaptability and scalability, representing a significant stride toward its clinical application. Additionally, our findings suggest a potential for identifying shared EEG signatures between MCI and dementia, contributing to the field's understanding of their common pathophysiological mechanisms.