Visión Electrónica
Approach to the diagnosis of cesarean delivery using bio- inspired models
Solo ficha. El repositorio marca este trabajo como «None»: el PDF no es público, así que aquí no hay texto para leer. Ver la ficha en el repositorio.
Resumen
In 2021, the cesarean section-related maternal mortality rate in Colombia was 46.4%. Efforts to reduce this rate have focused on monitoring maternal health, but the high volume of data and patient load complicate comprehensive symptom tracking. This study introduces a bio- inspired model for classifying cesarean deliveries using demographic information and electrohystereographic (EHG) biosignals from the mother-child dyad. The implemented classifiers include K-nearest neighbors (KNN), multilayer perceptron (MLP), support vector machines (SVM), and deep learning algorithms. For demographic data analysis, KNN achieved a sensitivity (S) of 100% and a specificity (ES) exceeding 80%, while SVM recorded an S of 75% and an ES of 83.3%. In EHG analysis, MLP demonstrated an S of 82.3% and an ES of 85.7%, followed by deep learning with an S of 72.8%. This model facilitates early detection of cesarean births by integrating maternal history and fetal behavior data.