Project DetailsIn this work we investigated and analyzed the connection of Blood pressure (BP) to biological signals Electrocardiogram (ECG) and Photoplethysmography (PPG). We used the MIMIC III database and obtained the relevant data. The signals are often noisy and unreliable; thus, an initial stage of noise filtering was needed. We defined several signal quality indices (SQI) and applied them on the data, filtering the unreliable data. Later, we divided our data to 30 second windows and developed feature extraction algorithms. We extracted several features, such as Pulse Arrival Time (PAT) and Heart Rate (HR). Using fluid mechanics principals, we developed a closed formula of BP as a function of the PPG signal and PAT feature. The BP formula included several calibration parameters, varying from one subject to another. Thus, we divided each subject in our group to a train set and test set, using the train set to train the closed formula, and the test set to check the quality of the estimation. Lastly, we optimized our estimation by defining an optimization energy function, utilizing the basic features of BP. We applied several tests and achieved a 16.74% relative improvement in comparison to the trivial estimation.
