A Shallow 1-D Convolution Neural Network for Fetal State Assessment Based on Cardiotocogram

Cardiotocography consists of fetal heart rate and uterine contraction signals that have been utilized for fetal well-being assessment. Researchers have applied several machine learning methods to improve the classification accuracy of the fetal state assessment. However, the proposed methods do not fulfill the required accuracy, as they have to address signal challenges such as missing value and external noise. Recently, convolutional neural networks have been brought to researchers' attention to cope with the challenges above in other machine learning applications. In this article, a new shallow architecture of 1-D convolution neural network is proposed to enhance fetal state assessment accuracy. This architecture has performed based on one convolution layer, resulting in computational complexity reduction. Besides, pooling operation that is a standard part of traditional CNN is not applied in this architecture to have more features in the classification phase. The performance of the proposed architecture is evaluated using five different clinical data sets. The results show that the proposed architecture is more efficient than traditional 1-D CNN and five implemented classifiers. The proposed architecture also achieves very competitive accuracy in the fetal state assessment compared to previous researches.