Mild Cognitive Impairment (MCI) is often seen as a precursor to Alzheimer’s disease (AD). Early identification of MCI is crucial for effective intervention, especially with emerging treatments. Recent research in the field of neurology has leveraged machine learning models to aid in clinical diagnoses of both AD and MCI.
A study in the journal Ophthalmology Science, conducted by Duke Health researchers, showcased a machine learning model’s ability to distinguish normal cognition from MCI using retinal images. This model analyzed 236 eyes of 129 controls and 154 eyes of 80 MCI patients, achieving a 79% sensitivity and 83% specificity. When applied to an independent test set, it yielded an impressive area under the curve (AUC) of 0.809 (95% CI, 0.681-0.937).
Dr. Sharon Fekrat, a senior author and professor at Duke University School of Medicine, discussed the iMIND study’s use of longitudinal research to explore the link between retinal changes and neurodegenerative diseases. She also addressed the challenges of implementing these machine learning models in real-world scenarios. Furthermore, Dr. Fekrat highlighted the potential of this technology to replace traditional cognitive assessments and its significance in early disease detection and prevention.