Related Experiment Video
Updated: Oct 14, 2025
08:52
Novel Object Recognition Test for the Investigation of Learning and Memory in Mice
Published on: August 30, 2017
74.8K
Discriminating memory disordered patients from controls using diffusion model parameters from recognition memory
Roger Ratcliff1, Douglas W Scharre2, Gail McKoon1
1Department of Psychology, Ohio State University.
Journal of Experimental Psychology. General
|November 4, 2021
Summary
Memory disordered patients show distinct cognitive patterns compared to controls. Machine learning accurately identifies Alzheimer's and mild cognitive impairment, aiding clinical diagnosis.
Area of Science:
- Cognitive Psychology
- Neuroscience
- Computational Psychiatry
Background:
- Memory disorders significantly impact daily life and cognitive function.
- Distinguishing between different types of memory disorders, such as mild cognitive impairment (MCI) and Alzheimer's disease (AD), is crucial for effective treatment.
- Traditional diagnostic methods can be supplemented with advanced analytical techniques.
Purpose of the Study:
- To investigate cognitive differences between memory disordered (MD) patients and healthy controls using diffusion model analysis.
- To evaluate the efficacy of machine learning techniques in discriminating between MD patients, MCI patients, and AD patients.
- To explore the potential of computational modeling as an adjunct to clinical diagnosis.
Main Methods:
- Diffusion model analysis was applied to item recognition and lexical decision tasks for 105 MD patients and 57 controls.
- Machine learning algorithms including linear discriminant analysis, logistic regression, and support vector machines were employed.
- Accuracy rates were calculated for classifying MD patients versus controls, and for differentiating within the MD group (AD vs. MCI).
Main Results:
- Diffusion model analysis revealed lower drift rates, wider boundaries, and longer non-decision times in MD patients compared to controls.
- Mild AD patients exhibited lower drift rates than mild MCI patients.
- Machine learning achieved approximately 83% accuracy in separating MD patients from controls, with higher accuracy for AD (90%) than MCI (68%).
Conclusions:
- Diffusion modeling provides insights into the underlying cognitive processes affected in memory disorders.
- Machine learning techniques demonstrate high potential for accurate classification of memory disorders, including differentiation between AD and MCI.
- These computational approaches may serve as valuable adjuncts to traditional clinical diagnostic procedures for memory-related conditions.
