300 uterine electrohysterogram (EHG) recordings categorized into 262 term and 38 preterm deliveries. Each record includes three-channel EHG signals sampled at 20 Hz along with clinical metadata such as maternal age and pregnancy duration.
Use Cases
- Develop classification models to distinguish between term and preterm pregnancies using the delivery outcome label
- Extract frequency-domain features from the 3-channel EHG signals to identify biomarkers for uterine activity
- Train regression models to predict the gestation age at delivery based on signal characteristics and maternal metadata
Strengths
- 300 total recordings consisting of 262 term and 38 preterm delivery cases
- Three-channel EHG signals recorded from the abdominal surface using four electrodes
- Clinical metadata including maternal age, parity, and gestation age at the time of recording
- Time-series data sampled at 20 Hz with a duration of approximately 20 minutes per record