Rethinking the residual approach: Leveraging machine learning to operationalize cognitive resilience in Alzheimer's...

Area of Science:

• Neuroscience
• Biostatistics
• Machine Learning

Background:

• Cognitive resilience (CR) is the ability to resist cognitive decline despite Alzheimer's disease (AD) neuropathology.
• Measuring CR is challenging as it's an unobservable, latent construct.
• The residual approach is commonly used but relies on strong, often unmet, assumptions.

Purpose of the Study:

• To critically evaluate the limitations of the residual approach for estimating CR.
• To propose and validate a novel, machine learning-based strategy for more accurate CR estimation.
• To overcome the inherent biases and inaccuracies of traditional CR measurement methods.

Main Methods:

• Critique of the assumptions underpinning the linear residual approach for CR estimation.
• Development of an alternative CR estimation strategy leveraging machine learning principles.
• Validation of the proposed method using simulated ground-truth data.

Main Results:

• The residual approach was found to introduce significant biases and errors in CR estimates.
• The proposed machine learning strategy demonstrated superior estimation accuracy on simulated data.
• The new method makes fewer assumptions, enhancing its applicability and reliability.

Conclusions:

• Traditional residual methods for estimating cognitive resilience are flawed and can yield erroneous results.
• Machine learning offers a more robust and accurate framework for measuring cognitive resilience.
• The developed approach provides a more reliable tool for understanding and quantifying cognitive resilience in AD research.