Related Experiment Video
Updated: May 5, 2026
Two-photon Imaging of Cellular Dynamics in the Mouse Spinal Cord
Published on: February 22, 2015
A multi-modal deep learning network for the classification of paramagnetic rim and remyelinated lesions in multiple
Federico Spagnolo1, Pedro M Gordaliza2, Aarushi Bhardwaj3
1Translational Imaging in Neurology (ThINK) Basel, Department of Biomedical Engineering, Faculty of Medicine, University Hospital Basel and University of Basel, Basel, Switzerland; Department of Neurology, University Hospital Basel, Basel, Switzerland Research Center for Clinical Neuroimmunology and Neuroscience Basel (RC2NB), University Hospital Basel and University of Basel, Basel, Switzerland; MedGIFT, Institute of Informatics, School of Management, HES-SO Valais-Wallis University of Applied Sciences and Arts Western Switzerland, Sierre, Switzerland.
Objectives:
Robust automated classification of paramagnetic rim lesions (PRLs) and remyelinated lesions based on iron and myelin content in people with multiple sclerosis (pwMS).
Materials And Methods:
In this prospective study (2018-2022), three-dimensional (3D) brain magnetic resonance imaging (MRI) from 180 pwMS (mean age: 47 ± 14 years, 108 females) were acquired at 3T including fluid-attenuated inversion recovery, magnetization-prepared 2 rapid acquisition gradient echoes (MP2RAGE), and T2*-weighted segmented echo-planar imaging. Post-processed quantitative susceptibility mapping (QSM) and filtered phase unwrapped (PU) images were generated. Ground truth was established through expert manual rating. Performance was assessed using nested cross-validation (CV) including outer test sets. Three neural network configurations were evaluated: (1) PRL classification with QSM and MP2RAGE, (2) PRL classification with PU and MP2RAGE, and (3) multiple lesion phenotype (MLP) classification with QSM and MP2RAGE.
Results:
For PRL classification, our network achieves a top mean validation F1 score of 0.737 ± 0.027 in the MP2RAGE-QSM configuration trained on PRLs, and a best test performance of 0.709 ± 0.040 when trained on MLP. The MP2RAGE-PU configuration yielded validation and test F1 scores of 0.733 ± 0.021 and 0.662 ± 0.037, respectively. The QSM-based MLP classification achieved a macro F1 test score of 0.728 ± 0.012.
Conclusion:
Deep learning can automate the classification of PRLs and QSM lesion phenotypes with high accuracy, potentially facilitating the detection of PRLs (included in the new diagnostic criteria) and remyelinated lesions, guiding personalized treatment decisions in pwMS.
Insights
This study developed a deep learning model for automated classification of paramagnetic rim lesions (PRLs) and remyelinated lesions in multiple sclerosis (MS) using MRI. The model shows high accuracy, aiding in diagnosis and personalized treatment for people with MS.
Area of Science:
- Neuroimaging
- Artificial Intelligence
- Medical Diagnostics
Background:
- Paramagnetic rim lesions (PRLs) are a key indicator in multiple sclerosis (MS) diagnosis and progression.
- Accurate classification of different lesion types, including remyelinated lesions, is crucial for understanding MS.
- Current manual classification methods are time-consuming and subjective.
Purpose of the Study:
- To develop and validate a robust automated classification system for PRLs and remyelinated lesions.
- To leverage deep learning models with quantitative MRI techniques for improved lesion characterization.
- To assess the performance of different neural network configurations for lesion classification in people with MS (pwMS).
Main Methods:
- Prospective study (2018-2022) involving 180 pwMS with 3T 3D brain MRI scans.
- Acquisition of fluid-attenuated inversion recovery, MP2RAGE, and T2*-weighted echo-planar imaging.
- Generation of quantitative susceptibility mapping (QSM) and filtered phase unwrapped (PU) images for analysis.
- Evaluation of three neural network configurations for PRL and multiple lesion phenotype (MLP) classification using nested cross-validation.
Main Results:
- The MP2RAGE-QSM configuration achieved a mean validation F1 score of 0.737 for PRL classification.
- The best test performance for PRL classification reached an F1 score of 0.709 when the model was trained on MLP.
- The QSM-based MLP classification demonstrated a macro F1 test score of 0.728.
Conclusions:
- Deep learning models can automate the classification of PRLs and QSM lesion phenotypes with high accuracy.
- Automated classification can aid in the detection of PRLs, supporting new diagnostic criteria for MS.
- This technology has the potential to guide personalized treatment decisions for pwMS.
