MINDSETS: Multi-omics Integration with Neuroimaging for Dementia Subtyping and Effective Temporal Study.

Area of Science:

• Neuroscience
• Medical Imaging
• Genetics

Background:

• Alzheimer's disease (AD) and vascular dementia (VaD) are common yet distinct cognitive disorders.
• Accurate differential diagnosis is crucial for effective treatment and improved patient outcomes.
• Current diagnostic methods often delay VaD diagnosis, hindering timely intervention.

Purpose of the Study:

• To develop and validate an innovative multi-omics approach for accurate differentiation between AD and VaD.
• To establish a new benchmark in diagnostic accuracy for dementia subtypes using integrated data.
• To introduce an interpretable model for enhanced clinical decision-making in dementia care.

Main Methods:

• Longitudinal MRI scans were segmented to extract advanced radiomics features.
• Radiomics features were synergistically integrated with clinical, cognitive, and genetic data.
• An ensemble learning approach was employed for classification.

Main Results:

• The proposed multi-omics model achieved a diagnostic accuracy of 89.25% in differentiating AD from VaD.
• The method demonstrated state-of-the-art classification accuracy on a large public dataset.
• An interpretable model was developed to support clinical decision-making.

Conclusions:

• The multi-omics approach offers a significant advancement in the accurate differential diagnosis of dementia subtypes.
• This methodology provides a nuanced understanding of dementia, paving the way for improved treatment strategies.
• Future research will focus on refining diagnostic capabilities and developing preventative measures for dementia progression.