Despite EEG is a clinically valuable signal for various brain-related disorders, EEG has been underutilized in the mental health practices due to labor-intensive denoising process and required expertise for the denoising, complex signal processing to get sensor or source level features, lack of biomarker relevant for various clinical process, and etc. But, recently developed iSyncBrain platform provides an efficient automated process targeted to various clinical circumstances even for telemedicine. The platform on the cloud comprised of AI-guided EEG denoising process (Kang, Jin, Keun Kim, & Kang, 2018) trained through 1800 normative EEG data collected during last 7 years, sex classified healthy normative QEEG database, standardized QEEG feature extraction process from adaptive mixture ICA (AMICA) dipole source information to sLORETA-based ROI connectivity, Normative library and group statistics for researchers (Kim, Kim, Kang, & Park, 2018;
D. Lee et al., 2018; Min et al., 2020)and series of QEEG discriminant biomarker has been implemented, such as early screening of Alzheimer dementia, prognosis of coma patients, brain age for development disorder and etc. (Han et al., 2021; Shim & Shin, 2020; Thapa et al., 2020) Biomarkers for stroke rehabilitation, Parkinson’s disease and early screening and intervention prediction for depression are under developing. (Baik et al., 2021; D. Lee et al., 2018; S. H. Lee, Ahn, Kim, Lee, & Lee, 2020; Min et al., 2020) In this lecture, a typical development process of machine learning model will be introduced using aMCI biomarker, a cloud based AI algorithm using 19 channel resting state EEG for early detection of prodromal stage of Alzheimer’s dementia. as an example. In the aMCI biomarker, the first QEEG discriminant functions on iSyncBrain platform, iSyncBrain iSB-M1 engine removes the noise signals and extracts various EEG features, which would be put into trained machine learning (ML) algorithms, then the probability score is calculated from 0 to 100. The scoring system consists of sequentially combined different ML algorithms, Alzheimer model, MCI model and amyloidopathy model. The clinical test with 429 participants, which is divided to two groups aMCI and Normal, finally shows 93.2% of sensitivity and 90.2 Specificity. Furthermore, possibilities of the methodology of the biomarker development, could be a catalyst for utilizing QEEG in various clinical circumstances in mental health area including neurology, psychiatry, psychology and etc, will be announced(Maestú et al., 2019). QEEQ experts in ISNR community can collaborate together for disseminate qEEG guided mental health methodology from a modeling of various mental diseases to an educational support to new comer in this field.
Presented by: Seung Wan Kang