mc-vivit
MC-ViViT: Multi-branch Classifier-ViViT to detect Mild Cognitive Impairment in older adults using facial videos
Sun, Jian, Dodge, Hiroko H., Mahoor, Mohammad H.
ADRD is technology has the potential to bolster the early detection of not only ranked as the sixth-leading cause of death in the MCI and AD. A 2019 study proposed that early detection of U.S., but the COVID-19 pandemic has increased the number AD would decrease healthcare costs, and improve quality of of deaths among those with AD (Alzheimer's_Association, life (Eikelboom et al., 2019). Studies have shown that MCI 2021). Symptoms of AD/ADRD usually begin with Mild can affect the patterns of speech, language, and face-to-face Cognitive Impairment (MCI) and include early onset memory communication in older adults (Liu et al., 2022a). There loss, cognitive decline, and impairments with verbal are several studies on using traditional Machine Learning language and visual/spatial perception. Approximately 12- (ML) approaches such as Decision Tree (Neelaveni 18% of people age 60 or older are living with MCI in the & Devasana, 2020; Davuluri1 & Rengaswamy, 2020) and U.S. Although older adults with MCI have the ability to K-means clustering (Davuluri1 & Rengaswamy, 2020) to independently perform most daily living activities such as detect AD with a human using collected demographic information eating, shopping, and bathing without help, the National (Neelaveni & Devasana, 2020) and MRI (Davuluri1 Institute on Aging (NIA) estimated that 10 to 20% of & Rengaswamy, 2020). Other researchers used Crossmodel subjects with MCI will develop dementia over a one-year Augmentation for automated detection of MCI from period (National_Institute_on_Aging, 2021).