JMIR Publications

ABSTRACT

Background:

Recent advances in cognitive digital assessment methodology, including high-frequency, ambulatory assessments, have shown promise to improve the detection of subtle cognitive changes. The use of computational modeling approaches may further improve the sensitivity of the digital cognitive assessments to detect subtle cognitive changes by capturing features that map onto core cognitive processes.

Objective:

We explored the validity of a brief, smartphone-based adaptation of a visual working memory task that has shown sensitivity for detecting preclinical Alzheimer’s risk. We aimed to optimize properties of the task for computational cognitive feature extraction with drift diffusion modeling.

Methods:

We analyzed data from a sample of 68 participants (69% women; 81% White; age range 24-80 years, mean 49 years, SD 14), who completed 60 trials for each of the 16 variations of a visual working memory binding task (the Color Shapes task) on smartphones, over an 8-day period. A drift diffusion model (DDM) was fit to the response time and accuracy data from the task. We experimentally manipulated three properties of the Color Shapes task (study time, probability of change, choice urgency) to test how these versions yield differences in key DDM parameters (drift rate, initial bias towards a response option, caution in decision making). We also evaluated how an additional task property, the test array size, impacted responses across all conditions. In terms of array size, we tested a ‘whole display’ of three shapes against a ‘single probe’ of one shape only.

Results:

The three the task property manipulations yielded the following results: (1) increasing the ratio of "different" responses was credibly associated with higher initial bias toward the “different” response (mean=0.06 for the whole display, mean=0.15 for single probe condition), (2) increasing the choice urgency during the test phase was credibly associated with decreased caution in decision-making in the single probe condition (mean=-0.04), but not in the whole display (mean= -0.01), and (3) contrary to expectation, longer study times did not yield credibly faster drift rate, but produced credibly lower ones for the whole display condition (mean=-0.28), and a null effect for the single probe condition (mean=0.01). In addition, as expected, we found that individual differences in drift rate were associated with age in both array sizes (r=-0.45 with BF=191), with older participants having lower drift rate. Older participants also showed higher caution (r=0.42, with BF=80.76) in the single probe condition.

Conclusions:

We identified a version of the Color Shapes task that is optimal for smartphone-based cognitive assessments in real-world settings when data to be analyzed through computational cognitive modeling. Our proposed approach can advance the development of tools for an efficient and effective early detection and monitoring of early risk for Alzheimer’s disease. Clinical Trial: Not applicable.