ABP Estimation Model

Developed a predictive system to estimate the ABP (Arterial Blood Pressure) using machine learning models, leveraging PPG (Photoplethysmogram) and ECG (Electrocardiogram) signals.

Built and deployed a Flask-based API for real-time predictions, allowing users to send PPG and ECG data via POST requests and receive ABP predictions.

Integrated a CNN-LSTM model for time-series data processing, improving the prediction accuracy of ABP values.

Implemented data preprocessing techniques using standard scaling to ensure optimal model performance.

Enhanced model efficiency by employing two scalers to handle input and output data normalization, achieving a reliable prediction system.

Deployed the API on a cloud server and tested it using ngrok for secure access.

Developed comprehensive logging and error-handling mechanisms to ensure robust API operations, minimizing prediction failures and issues.