Evaluation of ViewMind's ability to detect beta-amyloid and tau burden consistent with Alzheimer's disease in persons with MCI using Machine Learning, VR, and eye-tracking technology

PROJECT SUMMARY Alzheimer’s disease (AD) is the most common form of dementia which starts with gradual and wide-ranging cognitive and functional impairments mostly in people aged above 50 years. The initial phase of the disease is characterized by mild cognitive impairment (MCI). However, it is difficult to accurately distinguish between MCI caused by AD and MCI due to other causes using current neurophysiological assessment. MCI due to AD presents amyloid beta (Aβ) and tau depositions in the brain which can only be efficiently diagnosed by Positron Emission Tomography (PET) and cerebrospinal fluid (CSF)- procedures that are expensive, invasive, and complex4. The existing regular screening methods for early detection have limitations like limited scalability and accuracy. Since the failure of the standard of care for AD is largely due to multifactorial etiology and difficult-to- diagnose heterogenous clinical manifestations. Overall, there is a critical need for non-invasive, cost-effective, and rapid diagnostic tools that can accurately identify MCI due to AD at its earliest stages and measure functional impacts across cognitive domains, enabling timely interventions and the ability to monitor and quantify their effects. ViewMind is a digital health and artificial intelligence company that provides clinically validated solutions for precision diagnostics of neurocognitive disorders. ViewMind leverages ocular digital phenotyping, a head- mounted display (HMD), and machine learning (ML) to enable the early diagnosis of MCI due to AD, offering a non-invasive, cost-effective, and rapid method for precise measurement of multiple cognitive domains and their related brain regions. ViewMind’s eye-tracking technology evaluates eye movement responses across a series of cognitive tasks, collectively known as ViewMind Atlas AD exercises, within 15-20 minutes. The solution will incorporate multiple cognitive exercises tailored to assess specific domains and brain functions including the visual short term memory binding test (VSTMB). In this SBIR project, ViewMind will leverage data from the Bio- Hermes-002 study to develop and validate ML algorithms for both the early detection of MCI due to AD and the quantification of the functional impact on cognitive domains to measure the effects of early intervention. The specific aims of the current proposal are (i) Development of an ML algorithm using ViewMind Atlas AD exercises data to accurately classify participants into correct subject cohorts and (ii) Validation of the algorithm in classifying participants into correct subject cohorts in a double-blind study. After validating their technology in Phase I, ViewMind will validate the technology in clinical trials; followed by pursuing a Breakthrough Device Designation as its regulatory strategy.