Migraine Clinical Presentation Data for Automated Aura Classification

A research dataset of 400 patient records used to develop a machine learning classifier for distinguishing migraine with aura from migraine without aura. The dataset includes 23 clinical features such as demographic, temporal, and symptom characteristics. It was created by Sungjoon Hong and last updated on June 4, 2026.

Use Cases

• Training machine learning models for migraine subtype classification based on clinical features like visual disturbances and headache intensity.
• Analyzing feature importance for aura identification based on the reported discriminative capacity of symptoms.
• Evaluating model performance for clinical applications based on reported accuracy, F1-score, and sensitivity metrics.
• Supporting evidence-based therapy selection and cardiovascular risk stratification based on the classifier's output.

Strengths

• Dataset contains 400 patient records.
• Model development used 23 clinical features, with four key features (visual disturbances, headache intensity, dysphasia, age) accounting for 80% of discriminative capacity.
• The classifier achieved 87.50% accuracy and 90.91% F1-score on a held-out test set.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset is small (177.0 KB), indicating limited scope.

Provenance

Source

Sungjoon Hong via figshare.

Collection Method

Developed from patient records for a machine learning study.

Freshness

Last updated 2026-06-04 04:27:31; freshness should be verified.