EHG signal analysis for preterm labor detection


Angela Stephanya Caipe Gordillo and Jorge Armando Muñoz Rosero, Universidad de Nariño, San Juan de Pasto-Colombia 2017

The pre-term pregnancy occurs when labor occurs before 37 weeks of gestation, this fact is a major cause of mortality and morbidity in children, at present. Despite that there are several factors that indicate risk a pre-term delivery, it can be produced without the need of a symptom or indication factor. That is why several investigations aimed at solving this problem through a study of records of uterine electrical activity, known as electrohisterography (EHG), which represents a great hope when detecting a pre-term pregnancy.


Figure 1. Comparison of EHG records, a pre-term signal (a) and a term signal (b), the first 13940 points corresponding to 694 seconds of each record are plotted with a normalized voltage amplitude. 

By using computerized systems based on techniques of machine learning it is possible to determine the probability of pregnancy preterm from EHG records. Thereby was developed a methodology that includes signal preprocessing, features selecting techniques convex and non-convex, dimension reduction techniques, and finally supervised and unsupervised classification techniques.

For the classification in to features spaces process, a supervised and unsupervised analysis was used, using classifier based on Linear discriminant analysis (LDA), Quadratic discriminant analysis (QDA), classifier based on the non-parametric estimation as the classifier of Parzen, K Unsupervised analysis, among others.

According to the above, the classification algorithms are related to the results of a comparative study, which consists of evaluating the performance of combinations of characteristics techniques, dimension reduction techniques and classification techniques, that represent a balance between effectiveness, computational cost and easy interpretation of the physiological concept at the moment of performing the classification of EHG records.


Graphic Interface


In the Graphical Interface you can see the classification of the EHG signals with the different classifiers, in addition to the visualization of calculated meassures.




Publications

Thesis


Papers

Machine learning for the prediction of preterm pregnancy using EHG signals.
INCISCOS 2017 - International Conference on Information Systems and Computer Science  Quito - Ecuador.
Angela Stephanya Caipe Gordillo,Jorge Armando Muñoz Rosero, Diego Hernan Peluffo Ordoñez






Video Tutorial


About Us

 
Angela Caipe  Jorge Muñoz
 Electronic Engineer Electronic Engineer
 ascaipe@udenar.edu.co jorgem@udenar.edu.co





Comments