Advance Artificial Neural Network Classification Techniques Using EHG for Detecting Preterm Births

Abstract

-Worldwide the rate of preterm birth is increasing, which presents significant health, developmental and economic problems. Current methods for predicting preterm births at an early stage are inadequate. Yet, there has been increasing evidence that the analysis of uterine electrical signals, from the abdominal surface, could provide an independent and easy way to diagnose true labour and predict preterm delivery. This analysis provides a heavy focus on the use of advanced machine learning techniques and Electrohysterography (EHG) signal processing. Most EHG studies have focused on true labour detection, in the window of around seven days before labour. However, this paper focuses on using such EHG signals to detect preterm births. In achieving this, the study uses an open dataset containing 262 records for women who delivered at term and 38 who delivered prematurely. The synthetic minority oversampling technique is utilized to overcome the issue with imbalanced datasets to produce a dataset containing 262 term records and 262 preterm records. Six different artificial neural networks were used to detect term and preterm records. The results show that the best performing classifier was the LMNC with 96% sensitivity, 92% specificity, 95% AUC and 6% mean error.

Artificial Intelligence for Detecting Preterm Uterine Activity in Gynecology and Obstetric Care

“…The dataset used for this paper is the same as that used in [5], [9], [12], [13], with four features (root mean square, median frequency, peak frequency and sample entropy). The raw uterine EHG signal has been extracted from Physionet [7] using the Waveform Database (WFDB) toolbox.…”

“…In recent years, a number of researches have reported the use of a combination of machine learning classifiers, a better understanding of risk factors related to preterm birth and the use of Electrohysterography (EHG) signal processing has led to the introduction of several measures to improve the effectiveness of the essential treatments of pregnant women [5], [6]. EHG is used to measure electrical activity in the uterus, while machine learning algorithms are trained to distinguish between term and preterm EHG records through the detection of patterns in the data, while managing variance between subjects.…”

Training Neural Networks as Experimental Models: Classifying Biomedical Datasets for Sickle Cell Disease

“…In this paper, the previous work in the field of machine learning have been forced to check and predict the severe crises of SCD, rather than using optimal predication techniques to give correct amounts of hydroxyurea in modifying the disease phenotype [9]. Currently, there is no standardisation of disease modifying therapy management.…”

“…The back-propagation trained feed-forward neural network classifier (BPXNC): This type of neural networks is trained to set a number of input data through making adjustment of the complete weights [9]. In this context, the information t collected from imputes is fed forward to improve the weights between neurons.…”

“…Typically, this models has an input layer, a one hidden layer but without activation function, and output layer. The mapping procedure can be modified for RBNC through adjusted the weights values in the output layer [9].…”