This dataset offers insights into fetal health by analyzing features extracted from Cardiotocogram (CTG) exams. CTG exams are widely used in prenatal care to monitor fetal heart rate, movements, and uterine contractions, enabling early detection of potential health risks. The dataset has been classified by expert obstetricians into three categories: Normal, Suspect, and Pathological, providing valuable data for the development of predictive models aimed at reducing child and maternal mortality.

Fields

• Baseline Fetal Heart Rate (FHR): Baseline value of the fetal heart rate, measured in beats per minute.
• Accelerations: Frequency of heart rate accelerations per second.
• Fetal Movement: Number of fetal movements per second.
• Uterine Contractions: Number of uterine contractions per second.
• Light Decelerations: Frequency of mild decelerations in heart rate per second.
• Severe Decelerations: Frequency of severe decelerations in heart rate per second.
• Prolonged Decelerations: Number of prolonged heart rate decelerations per second.
• Abnormal Short-Term Variability: Percentage of time with abnormal short-term variability in heart rate.
• Mean Value of Short-Term Variability: Average value of short-term variability in heart rate.
• Percentage of Time with Abnormal Long-Term Variability: Proportion of time showing abnormal long-term variability.
• Fetal Health Classification: Labels indicating fetal health status:
  1. Normal: No evident risks to fetal health.
  2. Suspect: Potential risks, requiring further monitoring or intervention.
  3. Pathological: Critical condition necessitating immediate medical attention.

Usage

This dataset is well-suited for developing machine learning models for multiclass classification, particularly in the field of healthcare and prenatal diagnostics. Potential applications include:

• Predictive Modeling: Build machine learning models to classify fetal health into Normal, Suspect, or Pathological states.
• Risk Assessment: Identify key indicators of fetal health to inform early intervention strategies.
• Feature Analysis: Examine the relationship between CTG features and fetal health outcomes.
• Healthcare Optimization: Improve decision-making tools for healthcare professionals by integrating automated fetal health assessment.

Coverage

The dataset contains 2,126 records of CTG exams, reflecting a variety of fetal health states. The data is comprehensive and has been labeled with high reliability by obstetricians, making it a valuable resource for research into fetal health and maternal care.

License

The dataset is available under the CC0 (Public Domain) license, enabling unrestricted use for research, educational, and commercial purposes.

Who Can Use It

This dataset is ideal for:

• Data scientists and machine learning engineers focusing on healthcare.
• Researchers and medical professionals studying prenatal diagnostics.
• Students and educators exploring classification techniques in healthcare data.

How to Use

• Train machine learning models for fetal health classification to support early diagnosis.
• Analyze correlations between CTG features and health outcomes to identify key indicators.
• Create healthcare tools that integrate CTG data for real-time risk assessment.
• Design educational tools for obstetrics training using labeled CTG datasets.