Pulmonary Artery Pressure (PAP) Prediction Based On Three Physiological And Seven ECG Signal Features

Pulmonary arterial pressure (PAP), which is pro-duced by right ventricle of our heart by blood ejection among the pulmonary circulation, serves as a barrier to the right ventricular output. With every ventricular systole blood discharge, the vol-ume of blood in the pulmonary artery increase that extend artery wall. The blood keeps flowing into the pulmonary circulation, when the heart is relaxed. Regular PAP measurement is necessary to address the symptoms of Pulmonary hypertension (PH). PH is one kind of excess blood pressure for which the blood vessels of our lungs become narrowed, blocked or damaged. This reduces the flow of blood into the lungs and increases blood pressure within lungs arteries. Then our heart puts more effort to pump blood in our lungs which will gradually weaken and fail the heart muscle. To predict PAP from Electrocardiography (ECG) signal, we extracted seven ECG signal features using our developed algorithm and three other physiological features. Then, used them to train the Random Forest Regression algorithm. The final result displayed coefficient of regression (R
²
) of 0.98 for systolic and 0.83 for diastolic pressure. Mean absolute error, mean squared error, and root mean squared error is 2.8979, 15.3719, and 3.9207 respectively.