Needle Electromyography Signals for Patient Classification, 376 Signals

376 needle electromyography (nEMG) signals from 57 patients, used to validate a deep learning model for classifying neuromuscular conditions. The dataset was created by Ilhan Yoo and last updated in May 2026. It contains patient classification scores comparing model performance against six electromyographers.

Use Cases

• Training diagnostic models to classify patients as normal, myopathic, or neuropathic based on nEMG signals.
• Benchmarking AI model performance against human expert classification as described in the validation study.
• Analyzing signal features relevant for model interpretability and failure analysis as mentioned in the description.

Strengths

• Includes 376 nEMG signals from 57 patients, providing a basis for model validation.
• Performance metrics for the deep learning model (e.g., median accuracy 0.70) and physicians are provided for comparison.
• Data is shared under a permissive CC-BY-4.0 license.

Limitations

• Row count and column-level documentation are unknown, requiring manual inspection after download.
• The dataset is very small at 15.7 KB, indicating limited scope.
• Data may reflect bias inherent to the single tertiary university hospital source.

Provenance

Source

Tertiary university hospital database, author Ilhan Yoo.

Collection Method

Signals were collected and used to validate a deep learning diagnosis-aiding system via nested k-fold cross-validation.

Freshness

Last updated 2026-05-19 17:38:35.