JMIR Publications

ABSTRACT

Background:

Conventional memory tests for mild cognitive impairment (MCI) minimize the contextual cues of everyday life and may not capture process-level vulnerabilities relevant to daily functioning. Head-mounted virtual reality offers ecological immersion but can induce cybersickness and is difficult to deploy in outpatient settings. Tablet-administered 360-degree video offers a middle-ground that combines realistic-scene immersion with the accessibility of a familiar consumer device and a passive, exploratory serious-game framework for older adults.

Objective:

To provide a preliminary psychometric evaluation of two 360-degree video, tablet-based memory tasks, an episodic and a spatial memory task, administered to older adults with and without MCI, addressing convergent and divergent validity, age-adjusted group differences, discriminative accuracy, and the incremental contribution of digital process-level metrics over established neuropsychological measures.

Methods:

Eighty-two older adults (41 MCI; 41 cognitively healthy [CH]) were recruited at IRCCS Istituto Auxologico Italiano (Milan, Italy). Both tasks were delivered on an 11-inch iPad via the Mindscape platform, with task order counterbalanced. The episodic task yielded five outcomes (encoding, free recall, target retrieval, known- and unknown-distractor false alarms); the spatial task a single binary recognition outcome. Analyses included age-adjusted ANCOVAs, Pearson correlations with the Rivermead Behavioural Memory Test (RBMT) and executive-function measures, ROC analysis, and hierarchical binary logistic regression.

Results:

Seventy-six participants (35 MCI; 41 CH) completed the 360° session. The episodic task showed moderate convergent validity with the RBMT (r = .41–.52) and shared variance with executive-function measures, consistent with the demands of ecologically rich retrieval. After age adjustment, MCI participants performed worse on 360° free recall and produced more false alarms to semantically unrelated distractors (FDR-corrected); 360° free recall yielded a preliminary discriminative signal (AUC = .687, 95% CI .566–.807). Unknown-distractor false alarms, a process-level metric not directly indexed by conventional verbal tests, added incremental classification value beyond RBMT free recall after adjusting for age (OR = 2.05, p = .041). The spatial task showed a floor effect (overall accuracy 40.8%, above 25% chance but without group differences).

Conclusions:

The 360° episodic task shows preliminary psychometric properties consistent with a complementary, ecologically situated probe of memory processes - particularly susceptibility to false recognition of unrelated items, a process-level metric not directly captured by traditional verbal-episodic measures - rather than a stand-alone diagnostic tool. The spatial task contributes formative insights for serious-game design with older clinical samples, pointing to recalibration of distractor maps. Tablet-administered 360° video appears feasible across cognitive status and warrants larger, biomarker-characterized validation.