Aortic valve stenosis occurs when the aortic valve narrows, limiting blood flow from the heart to the artery and then to the rest of the body. It can cause cardiac failure in extreme circumstances. Identifying the illness in distant places can be challenging since it needs advanced technology and early stage diagnoses are difficult to get.
Researchers from the University of Kerala and the University of Nova Gorica in Slovenia published a method to discover valve failure using a complicated network analysis that is accurate, simple to use, and low-cost in the Journal of Applied Physics, published by AIP Publishing.
“Many rural health centers don’t have the necessary technology for analysing diseases like this,” said author MS Swapna, of the University of Nova Gorica and the University of Kerala, adding, “For our technique, we just need a stethoscope and a computer.”
The diagnostic technique is based on the noises made by the heart. While the mitral and tricuspid valves shut, the organ makes a “lub” noise, stops as ventricular relaxation occurs and blood flows in, and then makes a “dub” noise as the aortic and pulmonary valves close.
Swapna and her colleagues created a graph, or a complicated network of connected nodes, using heart sound data gathered over 10 minutes. The data was divided into portions, and each segment was represented on the graph by a node, or a single point. A line, or edge, was drawn between the two nodes if the sound in that piece of the data was comparable to another segment.
The network in a healthy heart exhibited two different groups of points, with many nodes remaining disconnected. A heart with aortic stenosis, on the other hand, has many more relationships and edges. “In the case of aortic stenosis, there is no separation between the lub and dub sound signals,” Swapna said.
The researchers examined the graphs using machine learning to identify individuals with and without illness, attaining a classification accuracy of 100%. Their approach evaluates the correlation of each point, making it more accurate than others, merely the intensity of the signal, and accomplishes so in less than 10 minutes. It might help with early-stage diagnosis.
So far, the procedure has only been evaluated with data and has not been tested in a clinical context. The writers are working on a mobile application that will be available worldwide. Their method might possibly be used to diagnose other diseases. “The proposed method can be extended to any type of heart sound signals, lung sound signals, or cough sound signals,” Swapna said.
