Combining Structural Magnetic Resonance Imaging and Visuospatial Tests to Classify Mild Cognitive Impairment

Area of Science:

• Neuroscience
• Medical Imaging
• Machine Learning

Background:

• Alzheimer's disease diagnosis benefits from combining multiple assessment methods.
• Identifying early stages, such as Mild Cognitive Impairment (MCI), is crucial.
• Machine learning (ML) shows promise for integrating diverse data types.

Purpose of the Study:

• To evaluate the efficacy of combining ML with neuroimaging and cognitive testing for early Alzheimer's detection.
• To identify key cognitive assessments for distinguishing MCI from healthy controls.

Main Methods:

• A pilot study used ML with structural MRI and cognitive tests.
• Participants included 11 healthy individuals and 11 with MCI.
• Cognitive assessment involved standardized and experimental visuospatial memory tests.

Main Results:

• The combined approach achieved 100% classification accuracy.
• Experimental visuospatial memory tests were particularly important for classification.
• High-level cognitive processing in these tests may act as a filter for early disease markers.

Conclusions:

• The integration of ML, neuroimaging, and advanced cognitive tests is effective for early Alzheimer's detection.
• Visuospatial memory tests are valuable tools for identifying prodromal Alzheimer's disease.
• This multimodal approach offers a promising strategy for diagnosing Mild Cognitive Impairment.