Dementia series part 5 — Dementia and Its Digital Biomarkers
Looxid Labs has developed the technology that analyzes EEG signals and eye-tracking data to assess cognitive performance. We are currently applying this technology to our in-home cognitive care solution, LUCY, designed to prevent dementia and spot the early signs of dementia in daily life. This dementia series aims to provide information on the basics of dementia and the latest digital technologies applied for early diagnosis, and last but not least, our technology and its use for early diagnosis of dementia through LUCY.
Written by Shannon H. Jeon
hyungsin.jeon@looxidlabs.com
In the article, Dementia series part 4 — Diagnostic Approaches to Dementia, we explored the biomarkers related to dementia and their importance in diagnosis and treatment processes. Biomarkers provide valuable evidence of the pathological changes associated with dementia; however, there are still many limitations to overcome. Currently, biomarkers can only be identified at hospitals using the high-cost and or invasive testing method. This type of testing procedure may hinder people from getting an early diagnosis of dementia since the testings are usually done only when symptoms are significant enough to suspect dementia. Therefore, further research on finding efficient dementia biomarkers is highly needed for early diagnosis and treatment. Various technologies and digital devices have been applied in dementia research to facilitate the development of efficient biomarkers, namely digital biomarkers.
Unlike the conventional biomarkers observing actual biological responses within the body, digital biomarkers are the indicators reflecting physiological or behavioral characteristics measurable outside the body. Using digital devices and their data, any phenomenon that captures a particular pathological state of a specific disease may be regarded as a digital biomarker or a candidate for further evaluation. Digital biomarker research includes easy-to-use wearable devices, non-invasive sensors, or advanced analysis technology that enhances the clinical value of existing physiological information. Examples are as follows.
1. Abnormal Eye Movement
When brain cells start to deteriorate, the consequences show in cognitive functions, such as memory and concentration, and the nervous system affecting various body functions. The eye movement function is one of the bodily functions affected by brain cell damage. When asked to aim our eyes to a particular spot, although it may seem like our eyes are fixated, it is expected to have tiny wobbles of the pupils lasting a few tens of milliseconds. However, when the brain is damaged, this kind of small, involuntary eye movement starts to get disturbed. Researchers use non-invasive tools such as high-speed cameras or eye-tracking devices to study the abnormalities in eye-movement dynamics.
The current dementia research scope of abnormal eye movement includes microsaccades, square wave jerks (SWJs), reaction time, gaze direction, etc. Microsaccades are small saccadic eye movements that occur when the gaze is fixed. Square wave jerks (SWJs) share similar features with microsaccades but have predominantly horizontal saccades moving away from the fixation. Reaction speed is measured by the latencies of eye response to a visual stimulus. The gaze direction is measured by the accuracy of saccadic eye movement when following a target. The results of eye movement studies related to cognitive decline show that the more severe the cognitive decline, the more microsaccades in the horizontal eye movements and gaze direction errors occurred. Researchers state that this phenomenon of cognitive impairment is due to the deterioration of the prefrontal cortex and the frontal eye fields that play a significant role in attention.
Other eye movement research studies cognitive impairment using anti-saccade tasks, in which participants are instructed to look in the opposite direction of a moving stimulus. Results indicate that the error of eye movement was more common in the cognitive impairment group. This may be due to the impairment in working memory and inhibition, the abilities needed to proceed with the task as instructed.
2. Decreased Language Ability and Abnormal Phonetic Patterns
Language and communication difficulties are common among patients with dementia. Even without significant decline in other cognitive functions, if the left brain cortex, the area responsible for language ability, is damaged, problems with word finding, naming, and pronunciation may appear. The current research has utilized various technologies to evaluate these symptoms, including the patient’s understanding and ability to read and write. The subtle linguistic patterns studied in dementia research include the abnormalities in vowel length, word stress, pauses between words, speech rhythm, etc.
A 2018 study of a total of 96 subjects — 48 healthy controls (HC); 16 patients with single domain amnestic mild cognitive impairment (aMCI); 16 patients with multiple domain MCI (mdMCI); and 16 patients with early dementia (eD) — reports the different linguistic patterns among groups. The speech analytics technology was used to differentiate the groups by their linguistic features, such as lexical, rhythmic, acoustic, and syntactic aspects. Notably, in the analysis of rhythm and complexity, not only the dementia group but also the two MCI groups could be distinguished from the healthy group. Another study also supports these results of speech analysis, reporting the classification accuracy as high as 87% (HC vs. MCI: 79%; HC vs. dementia: 87%; MCI vs. dementia: 80%).
3. Poor Game Performance
Scientists and researchers focus on developing serious games that could assess and improve cognitive function. Sea Hero Quest, released in 2016, is one of the serious games developed by neuroscientists. The game is designed to detect people at risk of Alzheimer’s disease (AD) by stimulating the brain regions responsible for specific cognitive functions. Players need to complete several tasks as a sea navigator, such as memorizing the map and navigating to the destination according to the map. The data generated during the game enables game developers and scientists to evaluate players’ spatial navigation ability. The appealing game story and graphics with a specific purpose have brought so much attention worldwide. Since its release, the game has exploded in popularity and has been played by 3.5 million people in over 200 countries, collecting vast amounts of information about users’ gameplay patterns.
The neuroscientists of Sea Hero Quest have found that the accuracy rate during the game performance, time spent on completing tasks, and in-game performance, such as routes game players took, are able to detect the people at risk of AD. According to the study of Proceedings of the National Academy of Sciences published in 2019, the game performance results of participants were able to discriminate between healthy aging individuals and those genetically at-risk of AD, with an accuracy rate of 71.3%. Since the participants were indistinguishable on neuropsychological tests, this finding is particularly meaningful, providing the steppingstone for individualized diagnostics through digital games.
Other serious games are based on a modification of the traditional neurocognitive tests, adopting automation technology. The gamified, automated assessments accurately collect the player’s performance speed and accuracy that will be utilized for cognitive function evaluation. For example, the N-back test, the Stroop test, and arithmetic test, typical cognitive tests, are adapted to digital games that can have better test effectiveness, resulting in more participants’ attention and eagerness to complete the tasks. Various ways to test cognitive function by using digital technology and gamification are constantly proposed.
Closing
Following the previous article, this article introduced a slightly different type of biomarkers: digital biomarkers. Whereas conventional biomarkers are indicators that relatively directly identify biological changes occurring within the body, digital biomarkers utilize digital technologies and tools to measure any information outside the body and find a relationship between that information and a specific disease. Various technologies and digital devices are already used in our daily lives. They hold great potential for early dementia detection. Diagnostic methods using digital biomarkers are relatively easy and low-cost, making it possible to apply to a large number of people and identify those at risk but undetectable in the current testing methods.
Looxid Labs is developing a solution called LUCY to help prompt cognitive management by measuring and monitoring cognitive performance using virtual reality and biosignal data, such as brainwaves and eye movement information. You can check your cognitive performance anytime, anywhere with LUCY. The patterns of brainwaves and eye movements during LUCY games, designed to target specific cognitive functions, may provide insights into your current cognitive ability. The next article will discuss how Looxid Labs’ EEG and eye movement information analysis technology can be used as an early diagnosis tool for dementia.
References
[1] Levy, N. K., Lavidor, M., & Vakil, E. (2018). Prosaccade and antisaccade paradigms in persons with Alzheimer’s disease: a meta-analytic review. Neuropsychology review, 28(1), 16–31.
[2] Kapoula, Z., Yang, Q., Otero-Millan, J., Xiao, S., Macknik, S. L., Lang, A., … & Martinez-Conde, S. (2014). Distinctive features of microsaccades in Alzheimer’s disease and in mild cognitive impairment. Age, 36(2), 535–543.
[3] Beltrami, D., Gagliardi, G., Rossini Favretti, R., Ghidoni, E., Tamburini, F., & Calzà, L. (2018). Speech analysis by natural language processing techniques: a possible tool for very early detection of cognitive decline?. Frontiers in aging neuroscience, 10, 369.
[4] König, A., Satt, A., Sorin, A., Hoory, R., Toledo-Ronen, O., Derreumaux, A., … & David, R. (2015). Automatic speech analysis for the assessment of patients with predementia and Alzheimer’s disease. Alzheimer’s & Dementia: Diagnosis, Assessment & Disease Monitoring, 1(1), 112–124.
[5] Coughlan, G., Coutrot, A., Khondoker, M., Minihane, A. M., Spiers, H., & Hornberger, M. (2019). Toward personalized cognitive diagnostics of at-genetic-risk Alzheimer’s disease. Proceedings of the National Academy of Sciences, 116(19), 9285–9292.
[6] Lumosity
