Related Experiment Video
Updated: Jun 19, 2026
10:46
A Method of Trigonometric Modelling of Seasonal Variation Demonstrated with Multiple Sclerosis Relapse Data
Published on: December 9, 2015
10.7K
A machine learning approach to determine the risk factors for fall in multiple sclerosis
Su Özgür1,2, Meryem Koçaslan Toran3,4, İsmail Toygar5
1Department of Biostatistics and Medical Informatics, Ege University Faculty of Medicine, Izmir, Türkiye.
BMC Medical Informatics and Decision Making
|July 31, 2024
Summary
Falls in people with Multiple Sclerosis (PwMS) are often caused by factors like disease impact and disability. Modifiable factors such as smoking and exercise habits can be targeted to reduce fall risks in PwMS.
Area of Science:
- Neurology
- Rehabilitation Medicine
- Data Science
Background:
- Falls are a significant concern for people with Multiple Sclerosis (PwMS), leading to injuries and functional decline.
- Identifying fall predictors is crucial for developing targeted interventions.
- Machine learning offers a novel approach to analyze complex fall-related factors in PwMS.
Purpose of the Study:
- To investigate the key factors contributing to falls in PwMS.
- To apply machine learning algorithms for predicting fall risk.
- To identify modifiable and non-modifiable risk factors for falls in the Multiple Sclerosis population.
Main Methods:
- A cross-sectional study involving 253 PwMS was conducted.
- Data collected using sociodemographic forms and validated scales (FES-I, BBS, FSS, EDSS, MSIS-29, T25-FW).
- XGBoost gradient-boosting algorithm used for predictive modeling of fall risk.
Main Results:
- The XGBoost model achieved an Area Under the Curve of 0.713.
- Significant risk factors for falls included MSIS-29 score, EDSS, marital status, education, disease duration, and age.
- Modifiable factors like smoking and regular exercise habits also emerged as predictors.
Conclusions:
- Smoking and regular exercise are identified as modifiable factors influencing falls in PwMS.
- Age and disease duration are non-modifiable risk factors that should guide risk identification.
- Interventions targeting MSIS-29 and EDSS scores, alongside education and financial support, can help prevent falls in PwMS.
