Related Experiment Video
Updated: Nov 4, 2025
07:15
Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
Published on: August 16, 2020
7.1K
Prediction of unenhanced lesion evolution in multiple sclerosis using radiomics-based models: a machine learning
Yuling Peng1, Yineng Zheng1, Zeyun Tan1
1Department of Radiology, the First Affiliated Hospital of Chongqing Medical University, No. 1 Youyi Road, Yuzhong District, Chongqing 400016, China.
Multiple Sclerosis and Related Disorders
|May 30, 2021
Summary
Machine learning models using radiomics can predict multiple sclerosis (MS) lesion evolution. The support vector machine (SVM) classifier with the ReliefF algorithm demonstrated the best performance in predicting lesion activity.
Area of Science:
- Radiology
- Machine Learning
- Neuroscience
Background:
- Multiple sclerosis (MS) lesion volume changes indicate disease activity and progression.
- Accurate prediction of lesion evolution is crucial for managing MS.
Purpose of the Study:
- To develop and optimize radiomics-based machine learning models for predicting the evolution of unenhanced MS lesions.
- To identify the optimal machine learning algorithm for this predictive task.
Main Methods:
- Prospective observation of 36 MS patients with 45 follow-up MRI scans.
- Lesion activity defined by >20% volume change on FLAIR images.
- Radiomic features extracted and analyzed using recursive feature elimination (RFE), ReliefF, and LASSO for feature selection, followed by logistic regression, random forest, and support vector machine (SVM) models.
Main Results:
- A total of 972 radiomic features were extracted, with 265 identified as robust.
- The SVM classifier combined with the ReliefF algorithm achieved the highest prediction performance.
- Achieved an accuracy of 0.827, sensitivity of 0.809, specificity of 0.841, precision of 0.921, NPV of 0.948, and an AUC of 0.857.
Conclusions:
- Radiomics-based machine learning models show significant potential for predicting the evolution of MS lesions.
- The developed models can aid in assessing disease progression and guiding treatment strategies.
