Identifying Predictors of Neck Disability in Patients with Cervical Pain Using Machine Learning Algorithms: A...

Area of Science:

• Biomedical Engineering
• Clinical Medicine
• Rehabilitation Science

Background:

• Neck pain is a common condition influenced by pain intensity, psychosocial factors, and physical function.
• Machine learning (ML) offers potential for classifying patients based on neck disability.
• Identifying key predictors is crucial for effective management of neck disability.

Purpose of the Study:

• To identify predictors of neck disability in patients with neck pain using clinical findings.
• To apply machine learning algorithms for classifying neck disability levels.
• To investigate the relationship between clinical variables and neck disability.

Main Methods:

• Ninety participants with chronic neck pain were assessed.
• Clinical data included pain intensity, neck disability index, cervical spine contour, and surface electromyography of axioscapular muscles.
• Machine learning models (MLP, LDA) were trained and validated using 10-fold cross-validation; MANCOVA was used for group comparisons.

Main Results:

• Multilayer Perceptron (MLP) achieved the highest adjusted R2 (0.768).
• Linear Discriminant Analysis (LDA) demonstrated the highest ROC area under the curve (0.91).
• Pain intensity emerged as the most significant predictor in both models, with a large effect size (0.568, p < 0.001).

Conclusions:

• Pain intensity is the primary predictor of neck disability in patients with cervical pain.
• Machine learning models effectively classify neck disability based on clinical features.
• Interventions targeting pain management are recommended for improving outcomes and reducing neck disability.