Prediction of Expanded Disability Status Scale in patients with MS using deep learning.

Area of Science:

• Neurology
• Medical Imaging
• Artificial Intelligence

Background:

• Multiple sclerosis (MS) is a chronic neurological disease causing significant patient disability.
• Accurate prediction of disease progression, particularly the Expanded Disability Status Scale (EDSS), is vital for effective patient management and treatment personalization.

Purpose of the Study:

• To develop a robust deep neural network (DNN) framework for predicting EDSS in MS patients using MRI data.
• To evaluate the model's performance in lesion segmentation and disability classification.

Main Methods:

• A deep neural network framework was developed and trained on MRI scans from MS patients.
• The model's performance was assessed using metrics such as Dice Coefficient, Jaccard Index, sensitivity, specificity, accuracy, precision, recall, and F1-Score.
• Ablation studies were conducted to evaluate the impact of different MRI sequences (T1-weighted, T2-weighted, FLAIR).

Main Results:

• The DNN achieved high accuracy in lesion segmentation (Dice Coefficient: 0.87) and disability classification (Accuracy: 91.2%, F1-Score: 0.885).
• Integrating T2-weighted and FLAIR images significantly improved prediction accuracy from 85.7% to 93.4%.
• The proposed model outperformed existing state-of-the-art methods in comparative analyses.

Conclusions:

• The developed DNN framework offers a reliable and accurate method for predicting MS disability progression using MRI.
• This technology has the potential to enhance personalized treatment strategies, enable early interventions, and improve the quality of life for MS patients.
• Further research should address data quality, sample size, and computational efficiency to facilitate real-world clinical applications.