Penza Researchers Develop Neural Network System for Heart Analysis
The system can detect signs of cardiomyopathy at an early stage.
Researchers at Penza State University and Penza State Technological University have developed an intelligent system designed to assist doctors in diagnosing cardiomyopathy, a disease of the heart muscle. The researchers say the condition is particularly difficult to identify at an early stage and is often mistaken for other disorders because its symptoms can be nonspecific.
The system translates established clinical criteria into mathematical decision equations that the Penza researchers derived for the first time and incorporated into the neural network's architecture. The software was trained on 75 examples from open databases. Its computer vision technology evaluates heart-wall thickness, the dimensions of the heart's chambers, ECG waveform characteristics, and other indicators.
The system has already completed state registration. Researchers plan to test it at Penza Clinical Hospital No. 6 in the near future.
