Related Experiment Video
Updated: Jul 17, 2025
12:18
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
7.6K
Decision trees to evaluate the risk of developing multiple sclerosis
Manuela Pasella1, Fabio Pisano1, Barbara Cannas1
1Department of Electrical and Electronic Engineering, University of Cagliari, Cagliari, Italy.
Frontiers in Neuroinformatics
|August 31, 2023
Summary
A new machine learning model uses immunogenetic markers to predict multiple sclerosis (MS) risk. This tool can help identify at-risk individuals and monitor family members of MS patients.
Area of Science:
- Neurology
- Immunogenetics
- Machine Learning
Background:
- Multiple sclerosis (MS) is a chronic neurological disease affecting the central nervous system, with uncertain etiology attributed to genetic and environmental factors.
- Current MS diagnosis involves clinical assessment, neuroimaging, and cerebrospinal fluid analysis, with no definitive cure available.
Purpose of the Study:
- To develop a predictive machine learning tool for assessing the risk of developing multiple sclerosis.
- To identify key immunogenetic risk markers associated with MS development.
Main Methods:
- A decision tree-based machine learning algorithm was developed.
- The algorithm integrated demographic factors (initially) and immunogenetic markers, including Human Leukocyte Antigen (HLA) class I alleles and Killer Immunoglobulin-like Receptor (KIR) genes.
- Demographic factors were excluded due to bias, focusing solely on immunogenetic markers.
Main Results:
- The study included 619 healthy controls and 299 MS patients from Sardinia.
- Excluding gender, the algorithm achieved 73.24% accuracy in identifying MS patients and 66.07% in identifying healthy individuals.
- The model demonstrated the predictive power of immunogenetic markers in MS risk assessment.
Conclusions:
- The developed machine learning system shows potential for clinical application.
- It can aid in monitoring relatives of MS patients and identifying at-risk individuals.
- Further research may refine predictive capabilities for early MS detection and intervention.
