Detecting Mild Cognitive Impairment via Digital Biomarkers of Cognitive Performance Found in Klondike Solitaire: A Machine-Learning Study

Abstract

Background: Mild cognitive impairment (MCI) is a condition that entails a slight yet noticeable decline in cognition that exceeds normal age-related changes. Older adults living with MCI have a higher chance of progressing to dementia, which warrants regular cognitive follow-up at memory clinics. However, due to time and resource constraints, this follow-up is conducted at separate moments in time with large intervals in between. Casual games, embedded into the daily life of older adults, may prove to be a less resource-intensive medium that yields continuous and rich data on a patient’s cognition. Objective: To explore whether digital biomarkers of cognitive performance, found in the casual card game Klondike Solitaire, can be used to train machine-learning models to discern games played by older adults living with MCI from their healthy counterparts. Methods: Digital biomarkers of cognitive performance were captured from 23 healthy older adults and 23 older adults living with MCI, each playing 3 games of Solitaire with 3 different deck shuffles. These 3 deck shuffles were identical for each participant. Using a supervised stratified, 5-fold, cross-validated, machine-learning procedure, 19 different models were trained and optimized for F1 score. Results: The 3 best performing models, an Extra Trees model, a Gradient Boosting model, and a Nu-Support Vector Model, had a cross-validated F1 training score on the validation set of ≥0.792. The F1 score and AUC of the test set were, respectively, >0.811 and >0.877 for each of these models. These results indicate psychometric properties comparative to common cognitive screening tests. Conclusion: The results suggest that commercial card games, not developed to address specific mental processes, may be used for measuring cognition. The digital biomarkers derived from Klondike Solitaire show promise and may prove useful to fill the current blind spot between consultations.

Digital Cognitive Biomarker for Mild Cognitive Impairments and Dementia: A Systematic Review

“…Two types of tasks for yielding digital biomarkers were identified: one is active tasks that require active engagement in the evaluation, such as computerized versions of traditional cognitive tests (e.g., [ 15 , 16 ]) or online games (e.g., [ 17 , 18 , 19 ]) for assessing a broad domain of cognitive functions, voluntary speech production for assessing language ability [ 20 ], or virtual reality techniques (e.g., [ 21 , 22 ]). Another type is passive tasks in which people were monitored without active input, such sleep quality, walking speed, phone usage, and daily activities [ 13 , 23 ].…”

Intercepting Dementia: Awareness and Innovation as Key Tools

“…Tech companies have been showing a growing interest toward the generation of “smart” devices for physiological monitoring and assessment of cognitive functions, both heavily affected in dementia-related disorders. The emerging technologies that hold great promise in the prognosis of neurodegenerative disorders cover a wide spectrum of applications, ranging from the development of wearable devices, such as fitness trackers, smartwatches, and mobile apps, to artificial intelligence platforms, VR, and game-based systems, purposely designed for early diagnosing, monitoring and treating dementia symptoms (Zygouris et al, 2017 ; Mandryk and Birk, 2019 ; Piau et al, 2019 ; Bloniecki et al, 2021 ; Gielis et al, 2021 ; Majnarić et al, 2021 ). Overall, digital tools provide a constant and extensive amount of data reflecting an individual’s health status.…”

“…Altogether, these observations suggest that cognitive testing through VR could give a complete and naturalistic setting to test for long-term cognitive changes. For instance, it has been proposed that digital biomarkers extracted from gameplay could reflect the user’s cognitive status (Gielis et al, 2021 ).…”

“…The introduction of digital tests, wearable devices, and VR systems offer the unprecedented opportunity to constantly track physiological parameters and cerebral functions, which can supply a substantial amount of data potentially underlying subtle anomalies occurring in the transitional phases of the pathology. Ultimately, allowing a faster and more accurate personal follow-up (Westerberg et al, 2010 ; Chester and Rudolph, 2011 ; Zygouris et al, 2017 ; Berg et al, 2018 ; Lucke et al, 2019 ; Mandryk and Birk, 2019 ; Piau et al, 2019 ; Zhavoronkov et al, 2019 ; Bloniecki et al, 2021 ; Gielis et al, 2021 ; Majnarić et al, 2021 ).…”

“…In line with the necessity of a more comprehensive approach, preventive methodologies are undergoing a fervent transformation, as demonstrated by an increasing focus on identifying innovative predictive tools. Novel digital technologies, such as wearable devices (smartwatches, fitness trackers) and mobile apps, have the ability to collect real-time physiological parameters, like sleep quality, blood oxygenation, and cardiac rate, as well as cognitive performance relevant for detecting preclinical dementia (Westerberg et al, 2010 ; Mancini et al, 2016 ; Piau et al, 2019 ; Gielis et al, 2021 ). Additionally, virtual reality (VR) and home-based systems are gradually attracting interest in the neuroscience field since their application allows the detection of minor cognitive changes and locomotor disabilities that remain otherwise unnoticed with the current methods (Zygouris et al, 2017 ; Piau et al, 2019 ).…”