Integrating Machine Learning into Myositis Research: a Systematic Review.

Area of Science:

• Rheumatology
• Autoimmune Diseases
• Computational Biology

Background:

• Idiopathic inflammatory myopathies (IIM) are autoimmune diseases causing muscle weakness and systemic effects.
• Clinical classification of IIM phenotypes impacts prognosis and pathophysiology.
• Machine learning (ML) is an emerging tool in IIM research.

Purpose of the Study:

• To systematically review the application of supervised ML models in idiopathic inflammatory myopathies.
• To evaluate the performance of ML models in specific IIM research contexts.

Main Methods:

• Systematic review of 23 original studies.
• Inclusion of supervised learning models: logistic regression (LR), random forest (RF), support vector machines (SVM), and convolutional neural networks (CNN).
• Performance assessment primarily using the area under the curve coupled with the receiver operating characteristic (AUC-ROC).

Main Results:

• Supervised ML models are being explored for applications in IIM research.
• ML applications include analysis of muscle biopsy transcriptome profiles, differential diagnosis via MRI and ultrasound.
• Performance metrics like AUC-ROC are key for evaluating ML model efficacy.

Conclusions:

• ML offers a complementary research tool for understanding IIM.
• Further research is needed to integrate ML into clinical practice for IIM management.
• ML shows potential in improving diagnosis and understanding disease mechanisms in IIM.