This project builds an intelligent system for classifying fetal health status using cardiotocographic (CTG) data. It predicts whether a fetus is in a Normal, Suspect, or Pathological state. The goal is to aid clinicians in making early and accurate decisions through data-driven insights.

• 📊 Exploratory Data Analysis (EDA) with visual insights
• 🧹 Data Preprocessing including feature scaling and handling class imbalance
• 🤖 Model Building: Tested multiple ML classifiers — Random Forest, XGBoost, SVM, etc.
• 📈 Evaluation Metrics: Accuracy, Precision, Recall, F1-score, ROC-AUC
• 🌐 Deployment: Real-time prediction app using Streamlit

The dataset used is based on CTG recordings, containing attributes such as:

• Baseline fetal heart rate
• Accelerations and decelerations
• Uterine contractions
• Short- and long-term variability

📌 Source: Dataset

Clone the repository:

git clone https://github.com/auspicie/Fetal_Health_Classification-ML.git  
cd Fetal_Health_Classification-ML  

Install dependencies:

pip install -r requirements.txt  

Run the Streamlit app locally:

streamlit run Fetal_health_classification_app.py  

The app provides a simple interface to input CTG values and predict fetal health status instantly.

Install dependencies

Fetal_Health_Classification-ML/
│
├── data/                       # Dataset files
├── notebooks/                  # EDA & Model training
├── app/                        # Streamlit app files
├── models/                     # Saved model and features
├── requirements.txt
└── Fetal_health_classification_app.py

This project is licensed under the MIT License.

Run the Streamlit app:

streamlit Fetal_health_classification_app.py

Contributions are welcome! If you’d like to improve this project or add new features:

fork → clone → make changes → submit a pull request  

Samsudeen A.
📧 auspicie@github
🔗 https://www.linkedin.com/in/samsudeen-bankole-b7b78554/

⭐ If you found this project useful, don’t forget to star the repo and share it!