Machine Learning and Symptom Patterns in Degenerative Cervical Myelopathy: Web-Based Survey Study.

Area of Science:

• Neurology
• Spinal Cord Injury Research
• Machine Learning Applications in Medicine

Background:

• Degenerative cervical myelopathy (DCM) is a progressive spinal cord injury due to spinal cord compression.
• DCM presents heterogeneously with sensorimotor, gait, and autonomic dysfunction, complicating diagnosis and understanding.
• Emerging evidence suggests DCM comprises distinct, yet undefined, patient subgroups.

Purpose of the Study:

• To investigate the utility of machine learning in identifying clinically meaningful patient clusters within DCM.
• To stratify DCM patients based solely on their reported clinical features.

Main Methods:

• A survey collected patient-reported clinical features, chief complaint, time to diagnosis, and demographics.
• K-means clustering, utilizing the Hartigan-Wong algorithm and Euclidean distance, grouped patients by clinical features.
• Clinical significance of identified clusters was assessed by comparing demographics, time to presentation, and disease severity.

Main Results:

• Three distinct patient clusters for DCM were identified through consensus.
• Clusters varied significantly in the total number of reported clinical features (p<.001).
• No significant differences were found between clusters regarding DCM severity or time since diagnosis.

Conclusions:

• Machine learning successfully delineated three DCM patient groups based on clinical presentation.
• These groups differ in symptom burden but not disease severity or duration, supporting DCM's heterogeneity.
• While machine learning shows promise for pattern recognition in DCM, high-quality datasets are crucial for deriving clinical benefit.