This cross-sectional pilot study evaluated the diagnostic accuracy of BraincheX — a self-administered digital cognitive test battery — for distinguishing early Alzheimer's disease (AD; n=25, MCI or mild dementia, amyloid-positive) from cognitively healthy controls (n=31, amyloid-negative) recruited from the LMU Munich memory clinic (October 2024–February 2025), using ROC analysis against the CERAD-Plus as a reference standard.
BraincheX total score achieved an AUC of 0.86 (95% CI: 0.76–0.95), sensitivity 89.7%, and specificity 68.0% at a cut-off of −1.51; it correlated strongly with the CERAD-Plus total score (r = 0.73, p < 0.001) and showed good internal consistency (Cronbach's α = 0.84), comparable to CERAD-Plus (AUC = 0.88, 95% CI: 0.76–0.97).
- Small sample (N=54 analyzed) with normative data derived from only 29 controls, limiting reliability of demographic adjustments. - Cut-off was derived and evaluated in the same sample with no internal validation (e.g., bootstrap), inflating apparent performance. - CERAD-Plus and BraincheX were administered an average of 4.9 months apart (range 0–12 months), introducing temporal confounding in convergent validity estimates.
BraincheX shows promise as a digital, non-verbal screening tool for early AD, with sensitivity comparable to CERAD-Plus and the advantage of automated scoring and reduced examiner burden. However, it should not replace a full clinical workup, and independent validation in larger, diverse cohorts is needed before routine clinical use.
Explore related topics
