DOI: 10.1002/alz.083236 ISSN: 1552-5260

Brief Continuous Recognition Task for Early Alzheimer Detection and Tracking

John Wesson Ashford
  • Psychiatry and Mental health
  • Cellular and Molecular Neuroscience
  • Geriatrics and Gerontology
  • Neurology (clinical)
  • Developmental Neuroscience
  • Health Policy
  • Epidemiology

Abstract

Background

Alzheimer’s Disease (AD) affects memory processing in the brain. An important aspect of AD assessment and study is memory testing, which can be done with excellent efficiency using a Continuous Recognition Test (CRT). A CRT was designed for online administration to provide 25 unique, complex, interesting images, which were initially presented for up to 3 seconds, and 25 repeated images, which required a response to indicate recognition (takes about 90 seconds). The images include 5 from each of 5 categories, from a set of 3,000 images, ensuring that a broad spectrum of memory is interrogated. The selected images are culturally neutral, and the instructions are provided in 120 languages, so the test is comparable across most countries, and performance is not affected by racial, cultural, or ethnic variations. The test is engaging and has been taken by many individuals over 100 times and by several individuals over 1,000 times. Two studies have compared this test with the Montreal Cognitive Assessment (MoCA) (in China and Netherlands) and found that this test performed at least as well.

Method

Data was examined from over a million CRT administrations from 3 sources: at the URL: www.memtrax.com, including over 18,000 unique users recruited by a French company (HAPPYneuron, INC), over 200,000 unique users from free access, and over 90,000 unique users from a free recruitment program (www.brainhealthregistry.org, part of the Alzheimer’s Disease Neuroimaging Initiative). Response time data was available for every image of each CRT, and the 3 populations were evaluated for HITs (correct recognitions), Correct Rejections (CRs, not responding to a new image), and response time to HITs (RT), including average RTs.

Result

Analyses indicated that 97% of individuals, across sites who scored better than chance (65% correct), performed better than 80% correct. Percent HITs and CRs and RTs showed a dichotomy with HITs (recognitions) correlating with RT (an AD dysfunction), while a separate group had fewer correct rejections (response disinhibition) and RT inconsistency, presumably related to fronto‐temporal dementia (clinical observation). Statistical analysis provides precision of impairment estimation.

Conclusion

A CRT can be used for early AD screening, memory assessment, and progression tracking.

More from our Archive