Naturalistic Driving Detects Mild Cognitive Impairment
Researchers at Arizona State University and collaborators in a 2026 JMIR Med Inform paper develop deep learning models that use naturalistic in-vehicle GPS, accelerometer, and gyroscope data to detect mild cognitive impairment (MCI). In a study of 22 participants, full-trip models achieved 78% accuracy and 77% AUC, outperforming turn-only inputs; the authors propose a frequency-based risk score for interpretable screening. Findings support scalable, noninvasive early cognitive monitoring.
Key Points
- 1Achieved 78% accuracy and 77% AUC using full-trip naturalistic driving data (N=22)
- 2Demonstrates full-trip data capture episodic MCI better than turn-only inputs
- 3Suggests frequency-based risk score enables interpretable, scalable screening in clinical and community settings
Scoring Rationale
Solid peer-reviewed application using naturalistic driving data, demonstrating practical accuracy, but small sample size and limited cohort constrain generalizability.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
