DIAGNOSIS OF FETAL HEART DEFECTS USING NEURAL NETWORKS

Introduction. This research addresses the critical exploration of fetal cardiac pathologies in pregnant women, aiming to enhance early detection methods through the utilization of sound data, specifically collected via microphones. This area warrants further exploration due to the urgent demand for more efficient methods to detect congenital heart problems in the womb, which would enable prompt treatment and better results.

Materials and Methods. Our study employs a novel approach by leveraging data from sensors, particularly microphones, to capture and analyze fetal heart sounds. The methodology integrates advanced neural networks to process and interpret the acoustic signals, seeking patterns indicative of various cardiac abnormalities. This innovative combination of sound processing technology and artificial intelligence offers a promising avenue for non-invasive prenatal diagnostics.

Results. Preliminary findings reveal the potential of the proposed methodology in detecting and classifying a spectrum of fetal cardiac pathologies. The neural network-based analysis demonstrates encouraging accuracy rates in identifying specific abnormalities, showcasing the feasibility of this approach as a reliable diagnostic tool. The results underscore the significance of acoustic data in unraveling intricate details of fetal cardiac health during pregnancy.

Conclusion. Our study underscores the transformative impact of sound processing technology and neural networks in prenatal diagnost...

Ряднов Н.С.