AI-Driven Eye-Tracking for Alzheimer's Disease Prediction
Author Information
Author(s): Li Qinjie, Yan Jiaxin, Ye Jianfeng, Lv Hao, Zhang Xiaochen, Tu Zhilan, Li Yunxia, Guo Qihao
Primary Institution: Shanghai Jiao Tong University Affiliated Sixth People’s Hospital
Hypothesis
Can an AI-driven eye-tracking tool on mobile devices effectively assess Alzheimer's disease?
Conclusion
The AI-driven eye-tracking technology can reliably identify differences in eye movement abnormalities in Alzheimer's disease.
Supporting Evidence
- The model showed an AUC of 0.85 for identifying Alzheimer's disease from normal controls.
- In the validation set, the model achieved an AUC of 0.91.
- Adjusted accuracy and central duration were significant predictors of Alzheimer's disease.
- Participants with Alzheimer's disease had significantly lower cognitive scores compared to normal controls.
- Eye movement parameters were significantly different between Alzheimer's patients and normal controls.
Takeaway
This study shows that we can use a smartphone to track eye movements and help find out if someone has Alzheimer's disease.
Methodology
Participants completed eye-movement tasks on a mobile device, and a logistic regression model was trained to classify Alzheimer's disease and normal controls.
Potential Biases
Potential biases due to participant selection and exclusion criteria.
Limitations
The study's eye movement paradigm needs validation in a larger sample size, and the app currently only supports Android devices.
Participant Demographics
273 participants, average age 70.46 years, 61.2% females.
Statistical Information
P-Value
p<0.001
Confidence Interval
95% CI: 0.80, 0.90
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website