An Enhanced Multiple Sclerosis Disease Diagnosis via an Ensemble Approach

Area of Science:

• Neuroscience
• Computational Biology
• Genetics

Background:

• Multiple Sclerosis (MS) is a significant neurological disease affecting over 2.8 million people globally.
• Current diagnostic methods can be improved with advanced computational techniques.
• Gene expression profiles offer a potential avenue for understanding and diagnosing MS.

Purpose of the Study:

• To develop an ensemble machine learning approach for accurate Multiple Sclerosis (MS) diagnosis.
• To address the challenge of class imbalance in gene expression datasets for MS.
• To identify key differentially expressed genes associated with MS.

Main Methods:

• A hierarchical ensemble classifier combining voting and boosting techniques was proposed.
• Heterogeneous voting was employed using Random Forest and Support Vector Machine as base learners.
• Feature selection methods, including BoostFS and DEGs, were utilized to identify relevant genes.

Main Results:

• The proposed ensemble approach achieved high diagnostic accuracy, reaching up to 93.5%.
• The method effectively handled class imbalance in gene expression data.
• The study successfully identified differentially expressed genes between MS patients and healthy individuals.

Conclusions:

• The developed ensemble approach is an efficient and accurate diagnostic tool for Multiple Sclerosis (MS).
• The identified differentially expressed genes provide insights into MS pathogenesis.
• This method advances the application of machine learning in neurological disease diagnosis.