Supported by Ministry of Science and Technology (MOST), Taiwan, the interdisciplinary research team, led by Professor Yi-Hung Liu at National Taiwan University of Science and Technology (NTUST), has successfully developed the world-first electroencephalography (EEG)-based computer-aided diagnosis (CADx) system for rapid screening of both dementia due to Alzheimer’s disease (AD) and its prodromal phase – the mild cognitive impairment (MCI), by integrating different domains’ technologies, including EEG signal processing, circuits and electronics, artificial intelligence (AI), cognitive neuroscience, and medical science. The developed EEG-based CADx system, has demonstrated a high accuracy and a high usability, and therefore, has great potential to provide an objective measure to assist MCI diagnosis, facilitating early intervention and reducing the risk of developing into Alzheimer’s dementia.
Dementia: One of the Leading Contributors to the Global Burden of Diseases.
According to the Alzheimer’s Disease International (ADI), There are currently estimated to be over 55 million people worldwide living with dementia. A new case of dementia arises somewhere in the world every 3 seconds. The number of people affected is projected to rise up to 139 million by 2050. In Taiwan, 291,961 elders suffer from this disease, occupying 7.7% of the order population (> 65 y/o) in 2020. It is estimated that the number of individuals with dementia will increase to 460,000 in 2031. Alzheimer’s disease (AD) is the most common and well-known form of dementia, accounting for 60–80% of all cases. Currently, there is still no effective treatment for the neurodegenerative disease such as AD.
Benefits to Early Diagnosis of MCI
Mild cognitive impairment (MCI) is a state between cognitive declines with normal aging and cognitive impairments caused by dementia due to AD. The MCI is a common condition in older population, and individuals with MCI are at higher risk to develop into AD in their later stage. Among individuals aged 65 years or older, the conversion rate from MCI to AD is 15% within 2 years and 32% within 5 years after the first diagnosis of MCI. There is increasing evidence that the risk of Alzheimer’s dementia can be greatly reduced by proper management of life style, status control of hypertension and diabetes (if any), good mental health care, pharmacological and/or non-pharmacological interventions at the MCI phase . However, MCI is usually underdiagnosed in community because older adults with MCI are often not aware of subtle declination of their cognitive function, which primarily prevents them from seeking for medical advices or even interventions. Therefore, a safe, objective, and easy-to-implement method for an accurate and efficient classification of individuals with MCI and heathy individuals is essential to promote early intervention of MCI.
Unique Solution to Accurate and Efficient Screen of MCI
Since 2017, Dr. Yi-Hung Liu and two other researchers, Dr. Chia-Fen Tsai (MD, Division of Geriatric Psychiatry, Taipei Veterans General Hospital) and Dr. Chien-Te Wu (International Research Center for Neurointelligence, The University of Tokyo), have delved into the research of EEG-based diagnosis for AD and MCI. The research team led by Dr. Liu has developed the nonlinearly multiple EEG feature decoding technique, and identified the most MCI-sensitive brain areas by using machine learning methods. An automatic EEG classification algorithm superior to the state-of-the-art methods was also proposed and embedded in the AI model, achieving a leading MCI detection accuracy of 90%. The related research outcomes have been published in different journals [2-4].
The EEG-based CADx system consists of three components, sensors (surface electrodes), EEG amplifier (analog signal acquisition and processing), and software (AI model). It has the advantages of convenience (less than 7 electrodes), safety (non-invasive), efficiency (only a 2-min EEG signal is required for analysis), and high accuracy (90%). Compared with other neuroimaging approaches, functional MRI for example, Dr. Liu’ technology is therefore more suitable for routine screen of MCI and AD. More importantly, so far, no similar EEG-based medical devices used for assisting diagnosis of both AD and MCI have got the FDA 510 510(k) and/or TFDA clearance for entering the market.
Linking a Startup and Top Medical Centers to Translate the Research Finding to a Real Product
To further translate the research finding into a smart medical device that can really be used in clinical practice, Dr. Liu initiated and co-funded a startup company, the Hipposcreen Neurotech Corp. (HNC), in 2019, which has received investment of more than 330 million US dollars. The EEG amplifier developed by HNC has already got the FDA 510(k) clearance and received TFDA approval in December 2020 and March 2021, respectively. At the same time, HNC has cooperated with major medical centers in Taiwan to establish the largest dementia and MCI EEG databank, including National Taiwan University Hospital, Taipei Veterans General Hospital, Chang Gung Memorial Hospital, and Tri-Service General Hospital (Taipei), aiming to collect EEG big data from at least 500 subjects. This goal is expected to achieve in the Q2/Q3 of 2022. Based on this databank, a more accurate and unbiased AI model for both AD and MCI prediction will be built. In other word, a total solution of the smart EEG system for screening of MCI, composed of the EEG amplifier, high-accuracy AI model, and a high-performance cloud-computing platform will be realized in the very near future.
1.Livingston, G.; Huntley, J.; Sommerlad, A.; Ames, D.; Ballard, C.; Banerjee, S.; Brayne, C.; Burns, A.; Cohen-Mansfield, J.; Cooper, C.; et al. Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet 2020, 396, 413–446
2.Y. T. Hsiao, C. T. Wu, C. F. Tsai, Y. H. Liu*, T. -T. Trinh, C. Y. Lee, ”EEG-based Classification between individuals with mild cognitive impairment and healthy controls using conformal kernel-based fuzzy support vector machine,” International Journal of Fuzzy Systems, 2021
3.T. -T. Trinh, C. F. Tsai, Y. T. Hsiao, C. Y. Lee, C. T. Wu*, Y. H. Liu*, ”Identifying individuals with mild cognitive impairment using working memory induced intra-subject variability of resting-state EEGs,” Frontiers in Computational Neuroscience, vol. 15, 2021.
4.Y. T. Hsiao, C. F. Tsai, C. T. Wu, T. -T. Trinh, C. Y. Lee, Y. H. Liu*, “MCI Detection using kernel eigen-relative-power features of EEG signals,” Actuators, vol. 10, 15 pages, 2021.