Early Detection of Pathology from Cardiac Pressure Estimation: Detection of Pathologies by Pressure Estimation using Image Signal Processing and Heart

Using a variety of techniques, such as image processing, simulation, and signal analysis, this study investigates novel approaches to predicting heart pressures linked to pulmonary artery problems. K-Nearest Neighbors (KNN) achieves 100% accuracy and Support Vector Machine (SVM) achieves 99.38% accuracy in classifying PAH severity levels. A MATLAB-based image processing technique measures pulmonary artery diameters from CT images to predict pulmonary artery hypertension (PAH), showing a significant correlation between diameter and cardiac pressure. In order to study the dynamics of pressure and velocity, A two-dimensional model of the right heart is created using COMSOL Multiphysics. This model reveals a pressure range of 825 Pa to 2.38 10⊃3; Pa at the initial state and notable variations of 1.71 10⊃3; Pa to 2.03 10⊃3; Pa during the cardiac cycle. Additionally, systolic Pulmonary artery pressure (SPAP) is estimated using spectral analysis of phonocardiographic signals from 200 healthy subjects. The results show elevated pressures that are inconsistent with expected clinical values for healthy individuals, with values of 65.81 mmHg and 64.93 mmHg. These results highlight the need for additional validation in a variety of patient demographics and demonstrate the potential and difficulties in accurately calculating heart pressures.